EA040692B1 - DIAGNOSIS SUPPORT PROGRAM - Google Patents

Abstract

A diagnostic support program is presented that is capable of displaying the movement of an area whose shape changes for each respiratory element, including all or part of an exhalation or inhalation. Provision is made for processing consisting of obtaining a plurality of frame images from a database storing images, processing consisting of determining a cycle of a breathing element that includes all or part of an exhalation or inhalation based on pixels in a specific area in each of the frame images, processing consisting of detecting a lung field based on the cycle of the determined breathing element, processing consisting of dividing the detected lung field into a plurality of block sections and calculating a change in the image in a block section in each of the frame images, processing consisting of a Fourier transform of the change in the image in each block section in each of the frame images, processing consisting of extracting a spectrum in a fixed band that includes a spectrum corresponding to the cycle of the breathing element from a spectrum obtained after the Fourier transform, processing consisting of performing an inverse Fourier transform of the spectrum extracted from the fixed band, and processing consisting of displaying each of the images on a display after performing the inverse Fourier transform.

Description

Field of invention

The invention relates to a method for analyzing images of the human body and displaying the results of the analysis.

Background of the invention

When a doctor is diagnosing a lung from chest dynamic images, it is important for the doctor to examine the chest dynamic time sequence images in which the subject is photographed in a state of natural breathing. Spirometer, which can easily obtain physiological data, RI (radioisotope) examination, simple X-ray photography, with which morphological data can be obtained, CT (computed tomography), etc. are known as lung function tests. However, it is not easy to obtain both physiological and morphological data effectively.

In recent years, an attempt has been made to obtain a method in which images of the dynamic state of the chest of a human body are photographed using a semiconductor image sensor such as an FPD (Flat Panel Detector) and the like used for diagnosis. For example, Non-Patent Document 1 discloses a technique for generating a difference image indicating a signal value difference between a plurality of frame images each constituting a dynamic state image to search for and display the maximum value of each signal value from the difference image.

In addition, Patent Document 1 discloses a technique for extracting a lung field region from each of frame images of a plurality of corresponding frame images indicating the dynamic state of the chest of a human body and dividing the lung field region into a plurality of small regions to perform analysis by linking the small divided regions to each other. between multiple image frames. In accordance with this technique, the magnitude of the feature is displayed, indicating the movement of a small divided area.

Documents of the prior art

Patent Documents.

Patent Document 1: Japanese Patent No. 5874636.

non-patent documents.

Non-Patent Document 1: Basic Imaging Properties of a Large Image Intensifier-TV Digital Chest Radiographic System Investigative Radiology: April 1987; 22:328-335.

The essence of the invention

Problems solved by the invention.

However, it is not easy for a doctor to determine the state of the disease only by displaying the maximum value of the difference between frames for each pixel of the dynamic state image, as in the technique described in Non-Patent Document 1. In addition, it is not enough to determine the state of the disease only by displaying the feature amount, as in the technique described in Patent Document 1. Therefore, it is desirable to display images according to the state of breathing and blood vessels of the lung. Those. it is desirable to determine the breathing state and the overall dynamic state of the blood vessels of the human body as a subject, and display images each of which shows the actual movement based on the trend of the waveform or respiration rate, heart and blood vessels or blood flow in the hilar region of the lung or their image.

The present invention has been made in view of such a situation and has the object of providing a diagnostic support program capable of displaying the movement of the area where the shape changes for each respiratory element, including all or part of exhalation or inhalation. More specifically, the objective is to generate diagnostic assist images by computing diagnostic aid numerical values by digitizing the concordance score or other nonconcordance score for the waveforms and frequency (Hz) already obtained for the new target data to be measured, and further by visualizing these numerical values.

The technical result provided by the invention is the ability to detect the state of breathing and the state of the blood vessels of the human body as a subject and display images each of which shows the actual movement based on the trend of changing characteristics of the state of breathing and the state of the blood vessels.

The object of the claimed invention is to generate images to assist in diagnosis, which makes it possible to improve the displays, which include technical aspects, including images and characteristics of the respiratory state and the state of the blood vessels of the human body and their trends.

Problem solving tool.

(1) To achieve the above objective, the present application takes the following steps. Namely, the diagnostic support program according to one aspect of the present invention is a diagnosis support program that analyzes human body images and displays the analysis results, the program causes the computer to execute a process, comprising processing to acquire a plurality of frame images from an image storage database; processing

- 1 040692 determining at least one frequency of the breath element, including all or part of the exhalation or inspiration, based on pixels in a specific area in each of the frame images; lung field detection processing based on at least one frequency of the detected breath element; processing of dividing the detected lung field into a plurality of block areas and calculating a change in the image in the block area in each of the frame images; processing of performing a Fourier transform of a change in an image in each block-section in each of the frame images; a spectrum extraction processing in a fixed band including a spectrum corresponding to the at least one frequency of the determined breath element from the spectrum obtained after the Fourier transform; processing for performing an inverse Fourier transform of the spectrum extracted from the fixed band; and display processing of each of the images after performing the inverse Fourier transform on the display.

(2) In addition, the diagnostic support program according to one aspect of the present invention has the ability to cause the computer to execute a process further comprising a spectrum extraction processing in a fixed band including a noise frequency and including a spectrum corresponding to a frequency other than the chip frequency. breath obtained from each of the frame images, or the input frequency or bandwidth, from the spectrum obtained after the Fourier transform, by using a filter.

(3) In addition, the diagnostic support program according to one aspect of the present invention has the ability to cause the computer to execute a process further comprising an inter-frame image generation processing based on the frequency of the breath element and each of the frame images.

(4) In addition, the diagnostic support program according to one aspect of the present invention is a diagnosis support program that analyzes human body images and displays analysis results, the program causing the computer to execute a process including processing, obtaining a plurality of frame images from a database storing Images; processing for determining at least one frequency element of cardiovascular contractions extracted from the heart beat or blood vessel beat of the subject; processing for determining at least one frequency of a breath element including all or part of an exhalation or an inhalation based on pixels in a specific area in each of the frame images; lung field detection processing based on at least one frequency of the detected breath element; processing of dividing the detected lung field into a plurality of block areas and calculating a change in the image in the block area in each of the frame images; processing of performing a Fourier transform of a change in an image in each block-section in each of the frame images; a fixed-band spectrum extraction processing including a spectrum corresponding to the at least one cardiovascular element frequency from the spectrum obtained after the Fourier transform; processing for performing an inverse Fourier transform of the spectrum extracted from the fixed band; and display processing of each of the images after performing the inverse Fourier transform on the display.

(5) In addition, the diagnostic support program according to one aspect of the present invention is a diagnosis support program that analyzes human body images and displays analysis results, the program causing the computer to execute a process including processing to obtain a plurality of frame images from a database storing Images; processing for determining at least one frequency element of cardiovascular contractions extracted from the heart beat or blood vessel beat of the subject; lung field detection processing; processing splitting the detected lung field into a plurality of block areas and calculating a change in the image in the block area in each of the frame images; processing of performing a Fourier transform of a change in an image in each block-section in each of the frame images; a spectrum extraction processing in a fixed band including a spectrum corresponding to the at least one cardiovascular element frequency from the spectrum obtained after the Fourier transform; processing for performing an inverse Fourier transform of the spectrum extracted from the fixed band; and display processing of each of the images after performing the inverse Fourier transform on the display.

(6) In addition, the diagnostic support program according to one aspect of the present invention has the ability to cause the computer to execute a process further comprising spectrum extraction processing in a fixed band including a noise frequency and including a spectrum corresponding to a frequency other than the chip frequency. CV beats obtained from each of the frame images, or an input frequency or bandwidth from the spectrum obtained after the Fourier transform by using a filter.

(7) In addition, the diagnostic support program according to one aspect of the present invention has the ability to cause the computer to execute a process further comprising an inter-frame image generation processing based on the frequency of the detected cardiovascular beat element and each of the frame images.

- 2 040692 (8) In addition, the diagnostic support program according to one aspect of the present invention is a diagnostic support program that analyzes human body images and displays the analysis results, the program causing the computer to execute a process including processing of obtaining a plurality of image frames from a database , which stores images; processing for determining at least one frequency of the blood vessel beat element extracted from the blood vessel beat of the subject; a processing of dividing an analysis range set for each of the frame images into a plurality of block areas, and calculating a change in the image in the block area in each of the frame images; processing of performing a Fourier transform of a change in an image in each block-section in each of the frame images; a fixed-band spectrum extraction processing including a spectrum corresponding to the at least one blood vessel beat element frequency from the spectrum obtained after the Fourier transform; processing for performing an inverse Fourier transform of the spectrum extracted from the fixed band; and display processing of each of the images after performing the inverse Fourier transform on the display.

(9) In addition, the diagnostic support program according to one aspect of the present invention has the ability to cause the computer to execute a process further comprising a spectrum extraction processing in a fixed band including a noise frequency and including a spectrum corresponding to a frequency other than the chip frequency. of the blood vessel beat obtained from each of the frame images, or the input frequency or frequency band from the spectrum obtained after the Fourier transform by using a filter.

(10) In addition, the diagnostic support program according to one aspect of the present invention has the ability to cause the computer to execute a process further comprising an inter-frame image generation processing based on the frequency of the detected cardiovascular beat element and each of the frame images.

(11) In addition, the diagnostic support program according to one aspect of the present invention is a diagnosis support program that analyzes human body images and displays the analysis results, the program causing the computer to execute a process including processing to acquire a plurality of frame images from an image storage database. ; processing for determining at least one frequency of a breath element including all or part of an exhalation or an inhalation based on pixels in a specific area in each of the frame images; lung field and diaphragm detection processing based on said at least one frequency of the detected breath element; processing of dividing the detected lung field into a plurality of block-sections and calculating a pixel change index in the block-section in each of the frame images; extraction processing of only a block-area, for which the custom index is in a predetermined fixed range, using a custom index that is a ratio value of a pixel change index in the block-area to a change index of a dynamic zone associated with breathing; and display processing of each of the images including only the extracted block areas.

(12) In addition, the diagnostic support program according to one aspect of the invention has the ability to cause the computer to execute a process, further comprising processing to determine at least one element frequency of cardiovascular contractions extracted from the heart beats or beats of the blood vessels of the subject, or at least at least one frequency of the beat element of the blood vessels extracted from the beat of the blood vessels.

(13) In addition, the diagnostic support program according to one aspect of the present invention has the ability that the logarithmic values of the adjustable index are determined as a fixed range including 0.

(14) In addition, the diagnostic support program according to one aspect of the present invention has the ability to cause the computer to execute a process further comprising lung field detection processing in another frame using at least one Bezier curve on a lung field detected in a particular frame.

(15) In addition, the diagnostic support program according to one aspect of the present invention has the ability that internal reference points are selected in the detected lung field, and the lung field is subdivided with curvilinear or straight lines passing through the internal reference points inside lung field.

(16) In addition, the diagnostic support program according to one aspect of the present invention has the ability that the interval between test points at the outer extent of the detected lung field and at the outer extent is made relatively large, and the interval between the internal reference points is made relatively large. small in accordance with the expansion index for each zone within the detected lung field.

(17) In addition, the diagnostic support program according to one aspect of the present invention has the ability that the interval between checkpoints in the detected lung field is made relatively large in the craniocaudal direction relative to

- 3 040692 similar to the human body or relatively large in the direction of a particular vector.

(18) In addition, the diagnostic support program according to one aspect of the present invention has the ability to cause the computer to execute a process further comprising processing to detect a lung field in another frame using at least one Bezier surface on a lung field detected in a particular frame.

(19) In addition, the diagnostic support program according to one aspect of the present invention has the ability to cause the computer to execute a process further comprising processing for detecting a range corresponding to an analysis range in another frame using at least one Bezier curve over a predetermined analysis range in specific frame.

(20) In addition, the diagnostic support program according to one aspect of the present invention has the ability to cause the computer to execute a process further comprising imaging processing of at least one of a lung field, blood vessels, or a heart using at least one Bezier curve.

(21) In addition, the diagnostic support program according to one aspect of the present invention is a diagnosis support program that analyzes human body images and displays the analysis results, the program causing the computer to execute a process including processing to acquire a plurality of frame images from an image storage database. ; analysis range determination processing using Bezier curves for all received frame images; and analysis target detection processing based on intensity change within the analysis range.

(22) In addition, the diagnostic support program according to one aspect of the present invention has the ability to cause the computer to execute a process further comprising feature calculation processing on the periphery of the detected analysis target.

(23) In addition, the diagnostic support program according to one aspect of the present invention has the ability that the iris is detected by calculating the difference in intensity for each of the continuous images to display an index indicating the location or shape of the dynamic zone associated with the detected diaphragm or breathing.

(24) In addition, the diagnostic support program according to one aspect of the present invention has the ability that an aperture that is not interrupted by other areas other than the aperture is displayed by changing the intensity threshold to interpolate the entire shape of the aperture.

(25) In addition, the diagnostic support program according to one aspect of the present invention has the ability to cause the computer to execute a process, further comprising processing to calculate said at least one breath element frequency from the location or shape of the detected diaphragm or the location or shape of the dynamic zone associated with breath.

(26) In addition, the diagnostic support program according to one aspect of the present invention has the ability to cause the computer to execute a process further comprising spatial normalization processing of the detected lung field or temporal normalization of the lung field using reconstruction.

(27) In addition, the diagnostic support program according to one aspect of the present invention has the ability that the breath element is corrected by changing the phase of the at least one frequency of the breath element or smoothing the waveform of the breath element.

(28) In addition, the diagnostic support program according to one aspect of the present invention has the ability that the waveform of any of the zones within the analysis range is determined, and the constituent elements of the frequency of the detected waveform are extracted to display an image corresponding to the constituent elements waveform frequency.

(29) In addition, the diagnostic support program according to one aspect of the present invention has the ability to detect the density in the analysis range, and from there remove the place where the density changes in a relatively large manner.

(30) In addition, the diagnostic support program according to one aspect of the present invention has the ability to cause the computer to execute a process further comprising processing for selecting said at least one frequency when performing an inverse Fourier transform based on a spectral structural relationship in an organ-specific cycling from a spectrum obtained after the Fourier transform.

(31) In addition, the diagnostic support program according to one aspect of the present invention has the ability that the X-ray photographing apparatus is controlled to adjust the X-ray irradiation interval according to the at least one breath element frequency.

- 4 040692 (32) In addition, the diagnostic support program according to one aspect of the present invention has the ability that after performing the inverse Fourier transform, only a block in which the amplitude value is relatively large is extracted and displayed.

(33) In addition, the diagnostic support program according to one aspect of the present invention has the ability to cause the computer to execute a process further comprising diaphragm or chest change amount calculation processing to calculate the change rate from the change amount by determining the diaphragm or chest after identifying the lung field. .

(34) In addition, the diagnostic support program according to one aspect of the present invention has the ability to cause the computer to execute a process further comprising multiplication of a specific spectrum by a coefficient, wherein extraction display is performed based on the specific spectrum multiplied by the coefficient.

(35) In addition, the diagnostic support program according to one aspect of the present invention has the ability that a digital filter is applied to the area that is the target of analysis to determine the frequency or waveform of the breath element after obtaining a plurality of frame images from the database. data that stores images.

(36) In addition, the diagnostic support program according to one aspect of the present invention has the ability that a plurality of frequencies of a breath element including all or part of an exhalation or an inhalation is determined based on pixels in a specific area in each of the images. frames for displaying on the display each of the images corresponding to the plurality of corresponding frequencies of the breath element.

(37) In addition, the diagnostic support program in accordance with one aspect of the present invention has the ability that, with respect to a specific range, for at least some one frame image, an image collected up to some fixed value is selected and displayed on the display. .

Invention effect.

According to one aspect of the present invention, it becomes possible to display the movement of the region where the shape changes for each respiratory element including all or part of an exhalation or an inhalation.

Brief description of the drawings

Fig. 1A is a diagram showing the general configuration of a diagnostic support system according to the present embodiment;

fig. 1B is a diagram showing an example of a lung region partitioning method;

fig. 1C is a diagram showing states in each of which the shape of the lung changes over time;

fig. 1D is a diagram showing states in each of which the shape of the lung changes over time;

fig. 2A is a diagram showing the change in intensity in a particular block and the result obtained by performing a Fourier analysis on it;

fig. 2B is a diagram showing the result of a Fourier transform obtained by extracting frequency components close to heartbeats and intensity changes of frequency components close to heartbeats obtained by performing an inverse Fourier transform on them;

fig. 2C is a diagram showing an example of extracting some fixed band from a spectrum obtained after a Fourier transform;

fig. 2D is a diagram schematically showing the rate of change for a lung;

fig. 2E is a diagram showing an example image of a lung field region sample;

fig. 2F is a diagram showing an example image of a lung field region sample;

fig. 2G is a diagram showing an example image of a lung field region sample;

fig. 2H is a diagram showing an example image of a lung field region sample;

fig. 3A is a diagram showing an example of imaging a lung field contour using both Bézier curves and straight lines and showing a lung field that is at its maximum state;

fig. 3B is a diagram showing an example of imaging a lung field contour using both a Bézier curve and a straight line and showing a lung field that is in a minimal state;

fig. 4A is a diagram in which before and after images of the lung field are superimposed between the previous frame and the next frame;

fig. 4B is a diagram showing a state in which a line showing a large gap is generated as a result of obtaining a difference between two original images according to FIG. 4A;

- 5 040692 fig. 4C is a diagram showing the difference value of the sum of intensity density at each location in the top-bottom direction of the image in FIG. 4B;

fig. 5 is a diagram showing the result obtained by approximating the relative position of the diaphragm by performing curve regression;

fig. 6A is a flowchart showing the overall flow of the respiratory function analysis according to the present embodiment;

fig. 6B is a diagram showing an example of an image displayed on the display;

fig. 6C is a diagram showing an example of an image displayed on the display;

fig. 7 is a flowchart showing the general scheme of a lung blood flow analysis according to the present embodiment;

fig. 8 is a flowchart showing a general scheme for analyzing another blood flow according to the present embodiment;

fig. 9 is a diagram showing an example of multiplying some fixed spectrum by a factor for the spectrum obtained after the Fourier transform;

fig. 10 is a diagram showing an example of lung field imaging using Bezier curves;

fig. 11 is a diagram showing an example in which the lung field is partitioned using Bezier curves;

fig. 12 is a diagram showing an example in which the lung field is partitioned using Bezier curves;

fig. 13 is a diagram showing an example of contrast between aortic blood flow amount waveforms and ventricular volume waveforms;

fig. 14 is a diagram showing an example of lung and near lung pixel values;

fig. 15 is a diagram schematically showing the general configuration of blood vessels in the human body.

Detailed description of embodiments

First, the basic concept of the present invention will be explained. According to the present invention, as far as the biological movement of breathing and blood vessels, the area and volume of the lung field or other in the human body, the relative movement captured in a repeating manner in a fixed cycle, a fixed repetition or a fixed movement (repeating sequence) along the axis time in all or part of the range is captured and measured as a wave. The measurement results of the wave (A) are the shape of the wave itself or (B) the intervals of the wave (frequency: Hz). These two concepts are collectively referred to as the underlying data.

There may be waves related in a similar way during the same period of time. For example, in the case of breathing, the following approximation can be conceptualized. (average of gross density change)®(chest change)®(diaphragm movement)®(lung function test)®(thoracoabdominal breathing sensor). Regarding the above (A) waveform itself, the concept of waveform customization is used to display images based on the above (custom waveform visualization). In addition, regarding the above (B) wave intervals (frequency: Hz), the concept of frequency adjustability is used to display images based on the above (tunable frequency imaging).

For example, in the case of the heart, as in the contrast example between the aortic flow magnitude waveform and the ventricular volume waveform shown in FIG. 13, the peak of aortic blood flow and the peak of ventricular volume, as well as the waveforms, do not coincide with each other. However, when setting the time interval to equal intervals as one cycle, for example, the time from t1 to t2, the time from t2 to t3, the time from t3 to t4 in FIG. 13; one cycle of aortic blood flow amount and one cycle of ventricular volume are repeated many times, and thus, with respect to each waveform, it can be said that the frequency is tuned. Looking at this waveform, one cycle is determined from the actual measurement value, as shown in FIG. 13 and the waveform can be predicted using the model waveform. Those. regarding how to get the waveform as the base data, it can actually be measured and obtained from the frequency (cycle), and the model waveform can be used for this. If a cycle (period) of an organ having a heart rate or the like is found, a waveform can be predicted, and thus it becomes possible to display a dynamic image of an organ based on this waveform by determining the waveform of aortic blood flow amount, ventricular volume and so on.

In addition, upon receiving a change in the density of respiration, the heart, the hilum of the lung, etc. a digital filter can be applied to it in advance so that the other element does not mix.

In addition, the concept of the breath element is used in accordance with the present invention. The breath element includes all or part of exhalation and inhalation. For example, one inhalation and exhalation may be considered to be broken down separately into one exhalation and one inhalation, and may also be considered limited to any of 0, 10, 20, 30, 40, 50, 60, 70, 80, 90 and 100% of one exhalation or one breath. In addition, estimates can also be made by extracting only a fixed fraction of each exhalation, such as only 10% of an exhalation. It is made possible to extract an image with higher accuracy by using any of these data or data obtained by using them in combination. In this case, calculations can be performed interactively many times.

Such an idea can likewise be applied not only to the breath element, but also to the circulatory element.

In the present description, when creating baseline data, the extractions of mutual components complement each other for feature values derived from one or more modalities (e.g., two or more of density in some fixed range, amount of change obtained from measuring volume, chest movement, diaphragm movement spirometry and thoracoabdominal breathing sensor) or multiple waveform measurement for the same breath cycle, etc., thereby improving accuracy. By doing so, it becomes possible to reduce artifacts and improve accuracy based on some fixed line prediction and the like. The text uses the term density, but in the image it refers to the amount of absorption of pixels in a particular area. For example, in the case of ST for air, bones, and water, -1000, 1000, and 0 are used, respectively.

In addition, fluctuations in axis, width, range, and frequency (Hz) due to the extraction of mutual components and width are evaluated. Those. by multiple superimposition, the frequency (Hz) axis reference is averaged, and each axis's optimal range, width, range, and frequency (Hz) are calculated using the dispersion. There is a case here where the frequency (noises) of a different behavior is extracted, and if its wave exists, the extent to which no wave is included is measured in a relative manner. Those. there is a case in which only a portion of the waveform is extracted from the entire waveform element.

The present patent specification distinguishes between the use of the terms density and intensity. As described above, the density means the amount of absorption and shows the high air permeability in the original XP images and the dynamic XP image, and air, water and bones should be displayed as -1000, 0 and 1000 respectively by digitizing the part showing high permeability as being white. . On the other hand, the intensity is the amount of relative change in density, such as the amount displayed through the transformation into the power of the signal and the width of the density by normalization. Those. intensity is the relative value of light-dark, the degree of emphasis, etc. When directly referring to the absorbance value in the XP image, it is represented as density or density change (A density). Then it is transformed, as described above, for the convenience of displaying the image, and is represented as intensity. For example, the intensity is given in the case where the color display is performed with a 256-stop gray scale from 0 to 255. This division of terminology is applicable to the case of XP or CT.

On the other hand, in the case of MRI, even though air, water, and bones are tried to be set as 1000, 0, and 1000 respectively, there is a situation in which the values vary greatly due to MRI pixel values, types of measuring machines, human physical condition , physique and measurement time; and how MRI signals are obtained, such as T1-derived images, etc., also varies depending on the institution and types of measuring machines, so it cannot be fixed. Accordingly, in the case of MRI, density determination cannot be applied, as in the case of XP and CT. Therefore, MRI deals with relative values from the initial extraction stage, thus being expressed as intensity from the outset. Then the signals for processing are also intensity.

From the above, it becomes possible to obtain basic data. For the base data described above, a new target to be measured is retrieved in some fixed width and range of the waveform and frequency (Hz) of the waveform of the base data described above. For example, extraction is performed in width, range, and waveform element only for breath extraction and blood vessel degree extraction. In addition, this waveform and frequency width (Hz) are relatively and collectively determined based on statistics by using a waveform element in another function, an artifact such as noise, etc., of a waveform of a different modality, which is considered to have different customizability, repeatability, performed many times, etc. Then adjustment and analysis is required (machine learning can also be applied). This is because as the width and range increase, another function element starts to be added, if they are too narrow, the function element itself is eliminated, and thus the range must be adjusted. For example, in the case of having data for multiple repeats, it is easy to determine the range, frequency, concordance width to measure, and so on.

Regarding the customizable concordance score.

In the present description, the image trend will be explained as an adjustable concordance score. For example, the lung field is detected and partitioned into a plurality of block-areas to calculate the average density (pixel value x) of these block-areas in each frame image. Then, the ratio (x') of the average pixel value of the blocks in each frame image to the change width (from 0 to 100%) from the minimum value of the average density (pixel value x) to the maximum value is calculated. On the other hand, by using the value of the ratio (x'/y') of the ratio (y') of the change (y) in the aperture of each frame image to the width of the change (from 0 to 100%) from the aperture's minimum location to the maximum location, only blocks-sections for which the value of the ratio (x'/y') is within a predetermined fixed range.

Here, the case where y'=x' or y=ax (a is the numerical value of the diaphragm amplitude or the coefficient of the numerical value of the density) means full concordance. However, this does not mean that only the case of full concordance indicates a meaningful value, a value having some fixed width should be extracted. Thus, in accordance with one aspect of the present embodiment, the fixed width is determined using logarithms (log), as described below. Those. when the ratio (%) is calculated in the case where y=x, full tuning concordance means log y'/x'=0. In addition, when extracting a value, the range of the tunable concordance score is narrow or (numerically narrow) range, for example, if it is determined that log y'/x'=-0.05~+0.05 in the range near 0, and the range of the tunable score concordance is a wide or (numerically wide) range, for example, if it is determined that log y'/x--0.5~+0.5 in the range around 0. That is, the logarithmic values of a custom score are defined as a fixed range including 0. The narrower this range, the higher the numerical value that is concordant within that range, and the custom score can be said to be higher.

When calculating the number by determining the value of this ratio for each pixel of the pixels for healthy people, a normal distribution is obtained, in which the case of complete concordance is taken as a peak. In contrast, for people with a disease, the distribution of this ratio value should be disturbed. In addition, as described above, the width determination method using logarithms is only one example, and the present invention is not limited thereto. Those. in the present invention, image extraction is performed as follows: (average of density change in the coarse range)®(change in chest)®(diaphragm movement)®(lung function check)®(thoracoabdominal breathing sensor movement)®(lung field area and volume ) and this also applies to methods other than the logarithm method. With such a method, it becomes possible to display a customized image.

In the case of blood vessels, with respect to the sequence of density changes (x (one waveform at the hilum)) occurring in response to the sequence of heart beats (y), there is a slight time delay (phase change), so it can be written as y=a'(x-t) (i.e. Y®X). In the case of complete concordance, since t=0, y=x or y=a'x. Similar to the iris case, when extracting a value, the range of the tunable concordance score is narrow or (numerically narrow) range, for example, if log y'/x is determined to be 0.05~+0.05 in the range near 0, and the range of the tunable concordance score is is a wide or (numerically wide) range, for example, if it is determined that log y'/x--0.5~+0.5 in the range around 0. The narrower this range, the higher the numerical value that is concordant within this range, and it can be said that the custom rate is higher.

In the case of other blood vessels, the part described above that reacts to the heart is excluded, and the density on the central side, which is built from the hilum of the lung, is used. Similarly, the case of peripheral blood vessels can be considered.

In addition, the present invention can also be applied to the circulatory system. For example, the change in heart density is directly related to the change in blood flow density between the hilar region and the peripheral lung field, and the change in the sequence of changes in heart density and the change in density of the hilar region undergo a kind of transformation and are transmitted as is. It seems that this occurs as a result of obtaining a small phase difference from the relationship between the change in the density of the heart and the change in the density of the hilar region. In addition, a change in the density of the hilar region of the lung, and the like. is related, as it is, to the change in lung field density due to blood flow, and thus it is possible to express adjustability by a unit (dependence of the concordance score in Y®X), which is reflected in the sameness of the score. In addition, it seems that the cervical blood vessel systems and the large blood vessel systems, for example, in the chest, abdomen, pelvis, four limbs, etc. associated directly or with a small phase with a change in density, built in the central cardiac blood vessels on their periphery. Then, when the density changes according to the background and propagates, it becomes possible to consider this as an adjustable indicator of concordance to propagate the density change situation.

In the present description, the amount of change in one image and the rate of change in one image can be given as total inhaled air volume ® total exhaled volume

- 8 040692 air. Then, in the case where the numerical value is found in a relative manner from the difference with the transparency of the surrounding air, in order to display as a relative value (standard density/differential signal intensity), when the change amount from the lung field density is set to 1, the change amount and the change rate may be extracted for each of (1) for different images for each image, image when 1 is set for each of them (general guess), (2) for different images for each of them, the ratio when the absolute value of the whole breath or the whole breath obtained by adding the density (change amount and change rate), or the absolute value of inhalation and exhalation is set to 1, and (3) for multiple exposures, the ratio obtained as the total sum of the density during each inhalation-exhalation (selected several times at moments time 10%), which is set to 1.

Also, in the case of 3D MR and the like. for the value (when it is set to 1 this time) obtained by summing the intensity (in the case of MR) or density (in the case of CT) of the entire breath, the difference of its intensity or density can be converted into peak inspiratory flow volume data (at rest or mandatory breath) and for this value the actual measurement value for respiration and respiration rate in each lung field area can be converted by finding the ratio of intensity or density when calculating 3Dx time, at least with MRI, CT, etc. Similarly, by a single input of cardiac output, it becomes possible that the distribution in the capillary phase of the flow in the lung field is an estimate of the transformation into a distribution of lung peripheral blood flow value or volume.

Thus the following is satisfied:

(amount of change on the image due to inhalation)x(their total number for inspiration)®(change on the image due to exhalation)x(their total number for exhalation)®(the volume of air inhaled during this time: the volume of natural breathing or forced breath)®(volume of exhaled air during this time: volume of natural breath or mandatory breath)®(amount of change in the volume of inhaled air or exhaled air of natural breath or mandatory breath during this time).

When subtracting only the amount of change from 10% or 20%, an estimate can be calculated by calculating (their total number)x(the amount of change over that time).

The extracted amount of change is rendered and extracted onto the image. This is an analysis of the respiratory function and an analysis of the blood vessels, as explained below. Then, an indicator of change in both the chest and the diaphragm is visualized. Here, there are cases where artifacts in the results are again eliminated, and feature extraction is performed by extracting new data from the extracted waveform, a waveform that becomes the first base data, a waveform of a different modality, etc., an environment waveform, and a waveform beyond repeatedly. The way to eliminate artifacts will be described below.

In addition, there is a case where the feature value is determined even if a change component extracted from a different one than those extracted as described above is excluded. For example, when detecting the movement of the gastrointestinal tract, an attempt is made to extract the movement of the gastrointestinal tract by eliminating the influence of breathing and the influence of blood vessels from the abdominal cavity.

Next, an embodiment of the present invention will be explained with reference to the drawings. Fig. 1A is a diagram showing the general configuration of the diagnostic support system according to the present embodiment. This diagnostic support system performs a specific function by causing the computer to execute a diagnostic support program. The main module 1 includes a respiratory function analysis unit 3, a lung blood flow analysis unit 5, another blood flow analysis unit 7, a Fourier analysis unit 9, a waveform analysis unit 10, and an imaging/digitizing unit 11. The main module 1 receives image data from the database 15 via the input interface 13. The database 15 stores, for example, images via DICOM (Digital Imaging and Communication in Medicine). The output image signal from the main unit 1 is displayed on the display 19 via the output interface 17. Next, the function of the main unit according to the present embodiment will be explained.

Analysis of the cycle of the element of respiration.

According to the present embodiment, the cycle of the breath element is analyzed based on the following indexes. The breath element means a concept that includes all or part of exhalation or inhalation as described above. Those. at least one frequency of the breath element is analyzed using at least one of density/intensity in a fixed area in the lung field, diaphragm movement, and chest movement. According to the at least one breath element frequency, one or more frequency spectra are indicated by the breath element, demonstrating the concept including the case of having a fixed bandwidth. It is assumed that the lung field is a collection of blocks, and a plurality of frequencies are extracted from each block, and thus, in the present embodiment, they

- 9 040692 are processed as a group of frequencies. In addition, as described earlier, the underlying data has both concepts of the waveform itself and the wave intervals (frequency: Hz), and thus it is also possible to perform processing as the waveform itself. In addition, a range consisting of the density/intensity of some fixed volume measured in a zone showing high X-ray transmissibility (besides, many kinds of modality such as others, including CT and MRI), data obtained by another measurement method can also be used. , such as spirogram, etc., and external information input.

In addition, the analysis results for each inhalation-exhalation are compared to analyze the trend from multiple pieces of data, and the accuracy of the data can also be improved.

In addition, it is also possible to correct the breath element by changing the phase of at least one frequency of the breath element or by smoothing the waveform of the breath element. In this case, the phase is matched to the waveform by using movement, eg (chest, diaphragm movement in addition)®(chest movement)®(density)®(fine lung function)®(chest sensor). In addition, the mean lung field density is monitored and the last change is fitted as a waveform by the method of squares for the wave to identify the wave. Here, in the case of chest density, etc. there are some cases where the change in lung density is assessed by assessing the density of the entire screen, since the biggest change is lung density. When constructing a wave, there is a case of actual movement and a case where the measured values generate a phase shift. In this case, a phase correction is made for the phase difference using the locations of the maximum and minimum values, the waveform as a whole, and the like.

Waveform analysis.

The frequency constituents of the waveform can be calculated from the waveform of the breath element. This produces an image with a customizable waveform as described above. Namely, a waveform in any of the zones in the analysis range is determined, and the frequency constituents of the determined waveform are extracted to display an image corresponding to the frequency constituents of the waveform.

Analysis of cardiovascular contractions and analysis of the beating of blood vessels.

According to the present embodiment, the cardiovascular contraction analysis and the blood vessel beat analysis are performed based on the following indices. Those. vascular beat is analyzed using the change in density/intensity of each zone by determining the heart/hila/major blood vessel location from measurements from other modalities such as an electrocardiogram and a tonometer, or from a lung contour. In addition, the change in density/intensity of the target zone can be analyzed by performing manual plotting on the image. Then, at least one frequency (waveform) of the cardiovascular contraction element obtained from the heart contraction or the beating of blood vessels is determined. In addition, it is desirable to improve the accuracy of the data by comparing the analysis results for each beat and trend across multiple data pieces. In addition, it becomes possible to improve the accuracy by performing the density/intensity extraction of each zone several times, as well as by performing the above relatively fixed range. In addition, there is also a way to enter the heart rate or frequency band.

Identification of the lung field.

The image is retrieved from a database (DICOM) and the lung contour is automatically detected using the breath element cycle analysis results as described above. A conventional technique for automatic lung contour detection can be used. For example, the technique disclosed in Japanese Public Unexamined Patent Application No. Sho 63-240832 or Japanese Public Unexamined Patent Application No. Hei 02-250180 can be used. The lung field is then divided into a plurality of block areas to calculate the change in each block area. In the present description, the size of the block area may be determined in accordance with the photographing speed. When the photographing speed is low, it is difficult to determine the corresponding area from the frame image following some frame image, and thus the block area is made large. On the other hand, when the photographing speed is high, the number of frame images per unit time is large, and thus tracking becomes possible even when the block area is small. In addition, the size of the block-section can be calculated according to which time reference is selected from the cycle of the breath element. In the present description, it is often necessary to correct the deviation of the lung field. In this case, thoracic motion, diaphragm motion, and the relationship of the location of blood vessels throughout the lung field are identified, and the relative position of the lung contour is then determined to generate relative estimates based on the motion. In addition, when the block area is too small, on

- 10 040692 The image flickers frequently. To suppress it, the block area must have a fixed size.

The lung field can be represented as coordinates of points and control points using at least one Bezier curve in the automatically detected region of the lung field. The lung field can then be represented by closed curves enclosed by at least one Bézier curve, i.e. sets of simple closed curves. Similarly, the goal of the analysis can also be represented by one or more simple closed curves.

The lung field in each frame allows the lung field to be detected in another frame using at least one Bezier curve on the lung field detected in a particular frame. For example, a method is provided for detecting corresponding two lung fields that are maximum and minimum to calculate a lung field in another frame using the two lung fields. In the present description, a variable called a change indicator is set for another frame. The change index is a value expressing the size of the lung field, ie. the state of inhalation-exhalation; and it can be calculated from the location of the aperture, from the average intensity of the entire image, and so on. It is also possible to calculate the measurement data of an external device such as a spirograph or the like and use the simulated change rate.

Thus, a variable called change rate can be set arbitrarily, and thus, for example, it can also be calculated assuming that the lung field changes with a fixed ratio (10, 20, 30%...). The change rate thus set includes errors, and thus there is a case where post-processing is performed using the result obtained by automatic/manual error elimination, the result obtained by least squares approximation, etc. P. Assuming that a linear deformation occurs to the maximum lung field and the minimum lung field, the lung field in each frame is calculated using the change rate of each frame using a linear transformation method, and the like.

In addition, the above-described processing may be applied in an arbitrary range for continuous frames. For example, when breathing, the lung field repeats the changes from maximum to minimum, but the maximum shape is not always constant in the actual dimension. For example, by applying the processing described above in each range from maximum to minimum and minimum to maximum, one can expect to calculate a lung field with high accuracy, instead of performing calculation by setting and calculating two lung fields that are maximum and minimum. In addition, in the present description, an explanation has been given as a specific example using the maximum and minimum, but the present invention is not limited to this, and due to an arbitrary range, it can be carried out, for example, with a location of 0 and 30%, or 30 and 100% in the middle. inhale-exhale.

In addition, although the accuracy is reduced, it is also possible to calculate the lung field in each frame from one lung field. For example, the vector of change in the lung field is determined by performing an assessment on the shape of the chest. Namely, a method for determining the change vector at each control point of the Bezier curve, but the present invention is not limited to this. The lung field is then calculated in each frame using the single lung field detected, the change vector, and the change rate in each frame. As for the calculation result, the accuracy can be further increased by performing automatic or manual correction. In addition, the method described above is also effective in the case of 3D. Those. even in the case of 3D, it is also possible to perform lung field detection processing in another frame using at least one Bezier surface on the lung field detected in a particular frame. Thus, it becomes possible to obtain an image of the lung field between frames.

Fig. 6C is a graph showing the cycle of the breath element. On the image of Fig. 6C shows a white vertical line, and it is a straight line (index) indicating the current position in the cycle of the breath element, and it moves to indicate the current position in the cycle of the breath element in accordance with the motion of the dynamic lung image shown in FIG. . 6B. It becomes possible to accurately determine the current position in the cycle of movement of the lung by indicating the current position in the cycle of the breath element. In addition, according to the present invention, it is possible not only to show the cycle of the breath element using a graph, but also to graph everything related to it, such as blood flow density, chest and diaphragm, the movement of which is associated with the movement of the lung.

In the case of holding the breath by the subject, there is a case where the frequency of the breath element cannot be determined. In this case, the Fourier analysis mentioned below is performed by using at least one frequency of the cardiovascular contraction element extracted from the patient's heartbeat or blood vessel beat. In this case, the division method of the block section mentioned below can be changed in accordance with the movement method of the dynamic zone associated with the heart, diaphragm, respiration, and the like.

Detection of the periphery and its evaluation.

- 11 040692

In accordance with the present invention, detection of the periphery of the lung is possible to evaluate the periphery. For example, after calculating the lung field using the above method, the location and shape of the periphery can be detected with high accuracy. A point is plotted at an arbitrary location within the computed lung field, and a line is drawn in all directions to evaluate the change in pixel values for each line. For example, as shown in FIG. 14, when calculating the pixel values along the segment S of the line that divides the lung, it is understood that the pixels vary greatly at its periphery, but the absolute value of the variation is different. For example, the detection accuracy of the periphery is improved by adjusting the threshold value at the time of detecting the periphery on the left side as well as the periphery on the right side. In addition, a variation characteristic of the pixel values for each area may also be used. As shown in FIG. 14, even though the difference at the edge of the area S2 and the area S3 is small, the edge of the area S2 and the area S3 can be determined from the variance of the pixel value variation. In this case, attention was paid to the dispersion, but the present invention is not limited to this.

In addition, in the same light, it becomes possible to detect the periphery of the analysis range of every organ, blood vessel, tumor, etc. other than the lung. For example, when a contrast agent is present in the blood vessels, the inside of the blood vessels can be clearly visualized, but it is difficult to accurately calculate the outside and thickness of the blood vessels. According to the present embodiment, the periphery can be accurately detected, and thus the features and shape of blood vessels within the analysis range can be calculated. Accordingly, it becomes possible to use for diagnosis by quantifying the thickness and outer circumference of blood vessels, which has traditionally been difficult to determine in the past.

Preparation of blocks-sections.

Next, a method for dividing the lung field into a plurality of block areas will be explained. Fig. 1B is a diagram showing the method of radially splitting the lung field from the hilum of the lung. For the lung, the side of the diaphragm moves more than the side of the apex of the lung, and thus the closer to the side of the diaphragm, the coarser the breakpoints can be. In addition, in FIG. 1B, lines (dashed lines) may be further drawn in the vertical direction to be divided into a plurality of rectangular (square) block sections. Accordingly, it becomes possible to more accurately represent the movement of the lung. In addition, it is also possible to split a lung using methods such as a lung transverse splitting method by plotting dots in the vertical direction of the lung, a vertical splitting method of the lung by plotting dots in the transverse direction of the lung, a method of plotting a lung apex tangent portion, and a tangent to diaphragm to define the point of intersection between the above tangents as a central point and split the lung using line segments drawn at some fixed angle to a straight line (for example, a vertical line) including this point, a method of splitting the lung with a plurality of planes perpendicular to the line a line connecting the part on the side of the diaphragm and the top of the lung (or gate of the lung), etc. These methods are also applicable to 3D stereoscopic images. In the case of 3D, each organ is captured as a space bounded by many curved planes or flat surfaces. The organ can also be broken into smaller parts. For example, when considering a 3D model of the right lung, processing can be done separately for the upper lobe, middle lobe, and lower lobe.

With regard to the lung field region, chest movement, diaphragm movement and the relationship of the location of blood vessels in the entire lung field are identified, and the relative position of the lung contour is determined, and judgment should be made in a relative manner based on the movement. Thus, in the invention of the present application, after the lung contour is automatically detected, the region indicated by the lung contour is divided into a plurality of block areas to average the image change value (pixel value) included in each block area. For example, as shown in FIG. 10, it is also possible to plot points on a Bezier curve at opposite edges of the lung, and connect them, and then a curve through each intermediate point can also be used. As a result, as shown in FIG. 1C, even though the shape of the lung changes over time, it becomes possible to monitor changes in the region of interest over time. On the other hand, Fig. 1D is a diagram showing changes over time in the case of partitioning into blocks, without taking into account the shape of the organ (lung in this case) that is the target of the analysis. With regard to the lung field area, the movement of the chest as described above, the movement of the diaphragm, and the relationship of the location of blood vessels in the entire lung field are identified, and the relative position of the lung contour is determined, and the evaluation should be performed in a relative manner based on the movement, but as shown in Fig. . 1D, when partitioned into blocks without lung field designation, due to changes in the lung over time, the region of interest extends beyond the lung field region, resulting in a blank image. Namely, since the movement of the diaphragm

- 12 040692 shows a strong movement by which the lung field is reduced, only the diaphragm and the total numerical values should not be corrected, but it is desirable to correct the lung field by including the component of the chest and many other elements in it. In addition, there is a way to enter the frequency of the breath element or frequency band. The area partition calculation can similarly be performed also for 3D.

Further, as shown in FIG. 11, it is also possible that in lung field A, using Bézier curves, internal control points are selected within the detected lung field, and the lung field is subdivided using curved or straight lines passing through the internal control points within the lung field.

Those. control points are provided not only on the lung field frame but also in the lung field region, the lung field region (A) is partitioned using these control points. In this case, as shown in FIG. 12, the interval between test points at the outer extent of the detected lung field and at the outer extent can be made relatively large (1), and the interval between the inner reference points can be made relatively small according to the expansion rate for each zone in the detected lung field (2 ). In addition, in lung field A, the interval between checkpoints can be made relatively large in the craniocaudal direction relative to the human body, or relatively large in the direction of a particular vector. The definition of this vector is arbitrary, but, for example, it can be defined towards the opposite side of the lung field from the pulmonary apex, and can be defined towards the opposite side of the lung field from the hilum, as shown in FIG. 1B. It is also possible to determine the vector in the direction corresponding to the structure of the lung. Thus, if the partitioning method of the lung field is unequal partitioning, it is possible to display an image considering the characteristics of each region. For example, the outer circumference of the lung field moves strongly, and its deviation is large, and thus the blocks are made large. On the other hand, the inner part of the lung field does not move much, and its deviation is also small, and thus the blocks are made small and precise. In addition, for example, the lung field on the side of the diaphragm moves strongly, and its deviation is large, and thus the blocks are made large. On the other hand, the lung field on the side of the head does not move much, and its deviation is also small, and thus the blocks are made small and precise. As a result, it becomes possible to improve display accuracy. This method is not limited to the lung field, and can be applied to the dynamic zone associated with breathing and the like. This method is applicable in the case when the lung in three dimensions is divided into separate lobes of the lung. In addition, it is also applicable when limited to bezier curves and areas on the lower side of the iris are displayed, such as the heart and other organs. In this case, it is also possible to split the area unevenly by defining the vector in a direction corresponding to the structure of the heart or other organs.

Artifacts are then removed to interpolate the image data. Those. when bone etc. is within the analysis range, it is represented as noise, and thus it is desirable to remove the noise with a noise cut filter. Conventionally, air and bone are set as -1000 and 1000 respectively for x-rays, and thus the high permeability part shows low pixel values and is displayed in black, and the low permeability part shows high pixel values and is displayed in white. For example, when displaying pixel values with 256 gradations, black becomes 0 and white becomes 255. In the lung field region, X-rays easily pass through the periphery where there are no blood vessels and bones, and thus the pixel values of an X-ray image become small and the x-ray becomes black. On the other hand, X-rays barely pass at places where blood vessels and bones are present, and thus the pixel values of an X-ray image become large and the X-ray image becomes white. The same can also be applied to other CTs and MRIs. In the present description, from the results obtained by analyzing the cycle of the breath element described above, it becomes possible to eliminate artifacts by interpolating the data using values in the same phase based on the waveform for each inhalation-exhalation. In addition, when it is found that the coordinates are different, the pixel value varies greatly, or the frequency and density become abnormally high, clipping is performed for them, and for the remaining acquired images, this can be easily used to calculate the frequency (Hz) of the diaphragm and adjust the lung field, for example , by identifying a continuously smooth waveform using the least squares method, etc. In addition, the image overlay provides (1) a case in which the acquired comparison images obtained by capturing each image before and after are superimposed with the coordinates themselves, and (2) a method of overlaying the relative position information on the base by relative stretching of the images after acquisition. each image before and after to the bottom. By using the methods described above, it becomes possible that the shape of the lung field is corrected and changes in the image in the block areas are corrected. At this time, artifacts are again eliminated in relation to the results and feature extraction via extraction is performed.

- 13 040692 of the new extraction waveform data, the waveform that becomes the first base data, the waveform of another modality, and the like, the periphery, and the waveform several times. In this case, the number of times may be one or more.

Here the reconstruction along the time axis will be explained. For example, when the inspiratory time at 15 fps is 2 s, 30+1 images are obtained. In this case, reconstruction for every 10% can be done by simply overlaying 3 images at a time. Here, for example, 0.1 s means 10% and in the case when only photographs at 0.07 and 0.12 s are obtained as an image, a reconstruction of 0.1 s is necessary. In this case, the intermediate value in the images before or after 10% (the average of both values) is given to perform the reconstruction. In addition, the time axis is taken, and the coefficient can be changed depending on the ratio of the time intervals. For example, when there is no photographed value for 0.1 s, and there is a difference in the time axis, and there are also photographed values at 0.07 and 0.12 s, the conversion is done as follows (value in 0.07 s)x2/5 +(value in 0.12s)x3/5 to perform reconstruction. In addition, from the amount of change in the mean inhalation-outlet and the aperture ratio, the relationship of location and change at this c-number is determined, and this value is taken as a coefficient to find the numerical ratio. In addition, it is desirable to include 0 to 100% of the maximum intensity differential projection and perform the calculation by providing ranges such as 1020% reconstruction, 10-40% reconstruction, etc. Thus, it is also possible to perform reconstruction at a one-inhalation-exit ratio for the unphotographed portion. In addition, according to the present invention, it is also possible to perform similar reconstruction not only for breathing, but also for blood flow, chest movement, diaphragm movement, and a number of movements associated with other movements. It is also possible to perform reconstruction for each block or for each pixel. In addition, it is desirable to include 0 to 100% of the maximum intensity differential projection and perform the calculation by providing ranges such as 10-20% reconstruction, 10-40% reconstruction, and so on.

In addition, the lung field is detected in the manner described above, and the detected lung field can be normalized. Those. the detected lung field is normalized spatially or normalized temporally using reconstruction. The size and shape of the lung field differ depending on the differences between the bodies of different people, but this allows you to display within a fixed range for normalization.

diaphragm and chest.

It also becomes possible to determine the movement of the diaphragm as well as the chest by identifying the lung field as described above. Those. function evaluation positioning can be performed on the image by calculating the chest curves as well as the diaphragm curves on the Xp (2D image) of the recognized diaphragm as a set of detailed coordinates, and by digitizing the rate of change and the amount of change in the mean or downward direction in the local part of the curve, and also the deformation index by means of curve fitting performed for the diaphragm as curves. In addition, function evaluation can be similarly performed on the image by calculating the curves in the periphery drawn in the chest, in addition to the diaphragm as a set of detailed coordinates, and by digitizing the change in the mean value and the curve. Motion function evaluation is performed by evaluating the above two change and change indicators as relatively related, and digitizing and visualizing (stationary areas are related in the same way, etc.) the different change indicators.

Next, a method for evaluating the diaphragm and chest will be explained. First, with regard to the diaphragm, a horizontal line showing the right and left sides, which is orthogonal to the axis (so-called median) of the body, is taken as the axis to represent the movement. The iris line is then straightened as a baseline. Those. the aperture line becomes a horizontal line. The movement of the diaphragm is then evaluated. In addition, the aperture line is lengthened and straightened to evaluate the orthogonal movement of the curve. Then, outside the chest, movement is assessed using the line that connects the corner of the chest and diaphragm to the apex of the lung as a baseline. The chest line is straightened as a baseline, i.e. the chest line is aligned with the lung apex - costophrenic angle line to assess movement. The chest line is extended to the baseline and straightened to assess the orthogonal movement of the curve. The curvature and radius of curvature of the chest and the line of the diaphragm are then assessed as described above. Then, the above-described change is calculated as the change amount and evaluated as the change indicator by differentiating this change amount.

Each of the FIGS. 6B and 6C is a diagram showing an example of an image displayed on a display. In FIG. 6B shows the movement of the left lung as a dynamic image. On the image of Fig. 6B shows a white horizontal line, but it is a straight line (index) indicating the location of the iris, and thus moves up and down following the movement of the iris when a dynamic image is displayed. Thus, an image diagnosis can be made by a doctor by detecting the iris and displaying an indicator indicating

- 14 040692 the location of the detected aperture, i.e. a white horizontal line indicating the location of the diaphragm. In addition, it becomes possible to diagnose the entire diaphragm or one region of the diaphragm from the right side or the left side, from the outer side or the inner side, and the like. in addition to part of the diaphragm by using lung field diaphragm line recognition to recognize all points. Similarly, similarly not only to the diaphragm, but also to the dynamic areas associated with breathing, such as the chest, etc., it becomes possible to determine the movement of the chest along the line at the location of the tangent line, and the like. and chest lines through lung field recognition. Thus, provided that the periphery is in motion, it becomes possible to detect the periphery by taking the difference for continuous images. For example, as a rule, the tumor is hard, and its environment is soft. Accordingly, the tumor moves slightly and its surroundings move actively, and thus the periphery of the tumor can be detected by difference.

In addition, in the case of 3D MRI, CT, etc., function evaluation location can be performed on the image by capturing the aperture surface as a single coordinate or a cubic curved surface; calculation of coordinates and curves as a set of detailed coordinates (a contour on the periphery of the diaphragm or a set of a flat surface and coordinates); and digitizing the rate of change and the amount of change in the average or downward direction in the local part of the curve in addition to the rate of deformation through the fitting performed for the diaphragm as a curved surface. In addition, evaluation of the function from the image can be performed by similarly calculating the curvilinear surface of the periphery drawn in the chest, excluding the diaphragm as a set of detailed coordinates, and digitizing the mean value and the rate of change of the curved surface. Motion function evaluation is performed by evaluating the above two change and change indicators as relatively related, and digitizing and visualizing (stationary areas are related in the same way, etc.) the different change indicators.

Fourier analysis.

Based on the cycle of the breath element and the beat cycle of the blood vessels, which are analyzed as described above, Fourier analysis is performed for the density/intensity value in each block area and the amount of its change. Fig. 2A is a diagram showing intensity change in a specific block and a result obtained by performing Fourier analysis on it. Fig. 2B is a diagram showing the result of a Fourier transform obtained by extracting frequency components close to heartbeats, and a change in intensity of frequency components close to heartbeats obtained by performing an inverse Fourier transform on them. For example, when the change in intensity in a particular block is subjected to a Fourier transform (Fourier analysis), the results shown in FIG. 2A. The results are then obtained as shown on the right side of FIG. 2B by extracting frequency components close to heartbeats from the frequency components shown in FIG. 2A. By performing an inverse Fourier transform on them, the change in intensity tuned to heartbeats can be obtained, as shown on the left side of FIG. 2B.

As shown in FIG. 9, it is also possible that a particular spectrum is multiplied by a factor and weighted. For example, it is possible to use this method to implement waveform customization. Those. As for how to select a frequency when performing the inverse Fourier transform, the inverse Fourier transform is performed after selecting a plurality of frequencies and multiplying by their ratio. For example, if one wishes to emphasize the spectrum showing the highest frequency, it becomes possible to double the spectral intensity. In this case, frequency continuity is not required. You can select spectra that are present at specific intervals.

In addition, it becomes possible to assess the position of the density of the heart by the shape (the area of the zone where the left lung is deepened by the shape of the extracted lung field) of the left lung (there are some cases of the right heart in case of transposition of the internal organs) and the location of the vertebral body and diaphragm. In this case, the ROI of the heart is taken to extract the density. When performing this extraction, an estimate is made with the relative values of the spectrum of respiration and blood flow using a coarse range. In addition, there are some cases where breathing rates and other artifacts are removed beforehand by performing filtering using the frequency band (Hz) (heartbeats 40-150 Hz, ® 0.67 to 2.5 Hz) generated by cardiovascular contractions and etc. In addition, since the position of the heart changes according to the state of breathing, the position of the heart relatively changes depending on the value of the shape of the chest, because the position of the chest varies, and more accurate extraction of cardiovascular beats is often performed, as well as extraction of the hilum of the lung, large blood vessels, and the like. In addition, similarly to the movement of the diaphragm, a method for calculating the frequency based on the contour of a heart that is in regular motion is provided.

Here, when performing the inverse Fourier transform of a spectrum including frequency components, the inverse Fourier transform can be performed by taking into account both the frequency elements

- 15 040692 t (respiratory rate and cardiovascular rate), determined by the density during respiration and blood flow, and spectrum bands (BPF: band pass filter can be used); or based on elements of any of them. At least one frequency when performing the inverse Fourier transform can be selected based on the spectral structural relationship with the inherent organ cycling from the spectrum obtained after the Fourier transform described above. In addition, it is also possible to identify the waveform of an area that becomes a particular organ or target of analysis from the structural relationships of a plurality of frequencies obtained after a Fourier transform (image preparation with custom waveform).

In addition, it is possible to use an AR method (autoregressive moving average model) so that calculations are performed quickly when performing the Fourier transform. According to the AR method, a method is provided for using the Yule-Walker equation or the Kalman filter in an autoregressive moving average model, and it is possible to supplement the calculations by using the Yule-Walker estimates, the PARCOR method, or the least squares method. Through this, it becomes possible to obtain an almost real-time image, facilitate calculations, and correct artifacts at a faster rate. It becomes possible to extract and display the nature of the image in each block-section using such Fourier analysis.

In addition, when performing this Fourier analysis, it is possible to use a digital filter method. Those. a digital filter by which the original waveform is subjected to arithmetic processing is applied by performing the Fourier transform of the original waveform and obtaining the parameters of each spectrum. In this case, the digital filter is used without performing the inverse Fourier transform.

In the present description, a fixed-band spectrum including a spectrum corresponding to a cycle of a breath element from a spectrum obtained after a Fourier transform can be extracted by Fourier transform of the change in an image in each block-area in each of the frame images.

Fig. 2C is a diagram showing an example of extracting some fixed band from a spectrum obtained after a Fourier transform. As regards the frequency f of the spectrum of the composite wave, the following relation holds:

1/f=1/f ₁ +1/f ₂ between fi (respiration component) and f ₂ (blood flow component) of each frequency becoming a composite source, and the following method can be used to extract the spectrum.

(1) A blood stream having a high spectral ratio is extracted.

(2) The spectrum is extracted by splitting in the middle of the peak of the spectrum corresponding to respiration/blood flow and the peak of a plurality of neighboring composite waves.

(3) The spectrum is extracted by splitting at the bottom of the spectrum peak corresponding to respiration/blood flow and the peak of a plurality of neighboring compound waves.

As described above, according to the present invention, this does not mean that a fixed BPF is used and a spectrum is extracted in a fixed band including the spectrum corresponding to the cycle of the breath element. In addition, in accordance with the invention of the present application, it is possible to extract a frequency (for example, additionally, the density/intensity in each zone and the heart beat element obtained from heart beats or blood vessel beats) different from the frequency of the breath element obtained from the frame image, from a spectrum obtained after a Fourier transform or a spectrum in a fixed band, including a spectrum (for example, a spectral model) corresponding to an externally entered frequency by the operator.

In the present description, in the case of a composite wave spectrum having only two components (respiration and blood flow), we have 50%+50% as a component, and in the case of three components, each is equivalent to one third. Thus, the spectrum of the composite wave can be calculated to some extent by what is the percentage of the spectrum of the breath component and what is the percentage of the spectrum of the blood flow component, and their spectral components and height. You can extract the spectrum for a large ratio (%). Those. the ratio of the blood flow/respiration component to the composite wave component is calculated, and a spectral value having a large blood flow/respiration component is calculated and extracted. In addition, with regard to aperture identification, there are some cases where only the spectrum or alias corresponding to the zone in which the frequency (Hz) becomes relatively constant, i.e. the area in which the change in frequency (Hz) is small is extracted from the data obtained by obtaining the frequencies of respiration and cardiac blood vessels. In addition, in the case of determining the spectral band when identifying the iris and the like. there are some cases of determining the spectral band in the range where the frequency change (Hz) is generated, and the area on its periphery. Often a constituent element of the waveform can be considered.

In addition, with regard to the spectrum when performing the inverse Fourier transform, it is possible to choose the case (simulation approach) of performing extraction from the simply simulated frequency and bandwidth using a zone that is high (at least one) and the case (area activity approach) of extraction high frequency or low frequency according to the spectral

- 16 040692 value based on the actual frequency of the band. In addition, when the heart rate is represented by the letter A and the lung frequency is represented by the letter B, A is subtracted from the entire frequency band to obtain B. In addition, with regard to the spectrum obtained from the Fourier transform, not only one place can be extracted, but also many places on the frequency axis.

As a result of this, it becomes possible to extract not only the case of an exact match to the cycle of the element of respiration or the cycle of the beating of blood vessels, but also the spectrum to be taken into account and contribute to the diagnosis of the image. In addition, it is known that respiration and heart beats are included in a specific frequency band. Accordingly, using, for example, a 0 ~ 0.5 Hz filter (respiratory rate 0 ~ 30 times/min) in case of breathing, and using, for example, a 0.6 ~ 2.5 Hz filter (heart rate/pulse 36 ~ 150 times/min) in the case of the circulatory system, it is possible to predetermine the respiratory rate and the circulatory system frequency using the above filter. This allows you to display the image at a customizable frequency. This is because there is a case where a change in the density of respiration (lung) is captured when a change in density of the heart is obtained, and a change in the density of the heart is captured when a change in density of the lung is obtained.

Visualization/digitization.

The results of the analysis described above are visualized and digitized. When performing imaging and digitizing in the present patent specification, a simulated lung is specified. When displaying a lung with a dynamic image, the relationship of locations changes, and thus making a relative determination is not an easy task. Therefore, the deviation of the relationship of locations is spatially unified and averaged. For example, the shape of the lung is matched to a pattern such as the shape of a fan wing and the like, and displayed in a state in which the shape is matched. Temporary unification is then performed using the reconstruction concept. For example, a 20% lung condition is extracted from multiple breaths and is defined as a 20% lung condition for one breath. Thus, a lung that is unified in space and time becomes a simulated lung. Accordingly, when comparing different patients with each other or comparing one patient in the present moment with the same patient in the past, relative determination is easily performed.

For example, as a standard implementation, the value is often displayed relative/logarithmically, having an average value of 1 of the density/intensity over the entire area of the measured lung field. In addition, since only the direction of blood flow is used, a change in a specific direction is often cut off. By doing so, it becomes possible to extract only the data in a meaningful way. After changing the analysis range, pseudo-colorization is performed using the lung field identification result. Those. the analysis result of each individual (patient) is fitted to a relative area according to a specific shape (minimum, maximum, mean, median) fitted to the phase.

In addition, deformation of a specific shape/phase is performed, making it possible to compare multiple analysis results. In addition, when preparing the simulated lung, the relative relationship of locations in the lung field is calculated using the results of the breath element cycle analysis as described above. In addition, the simulated lung is prepared using the line obtained by comprehensively averaging the lines of chest, density, diaphragm, etc. many patients. When preparing a simulated lung in the case of lung blood flow, the distance can be measured radially from the hilum of the lung to the tip of the lung. In addition, in the case of breathing, it is necessary to make a correction in accordance with the movement of the chest and diaphragm. In addition, the calculation can be performed in a composite manner, taking into account the distance from the top of the lung.

In addition, after performing the inverse Fourier transform, only a block in which the amplitude value is relatively large can be extracted and displayed. Those. in the Fourier transform for each block after performing the inverse Fourier transform, there are blocks in which the wave amplitude is large and blocks in which the wave amplitude is small. It is then also efficient that only the block where the amplitude is relatively large is extracted and rendered. In addition, after performing the inverse Fourier transform, the real part and the imaginary part of each numerical value can be used properly. For example, it is possible to reconstruct an image only from the real part; reconstruction of the image only by the imaginary part; and reconstructing the image from the absolute value of the real part and the imaginary part.

The simulated lung can be Fourier transformed. It is also possible to use a simulated lung when comparing images of multiple breaths and performing Fourier analysis to determine relative position. The relative position when performing a Fourier analysis can be made constant by fitting a set of received frames to a simulated lung, and fitting them to a simulated lung calculated according to heartbeats (e.g. heartbeats obtained from the hilar region of the lung)

- 17 040692 in the case of blood vessels. By obtaining a breathing state that becomes a baseline, it also becomes possible to stably obtain calculation results using a simulated lung. In addition, it becomes possible that a spatial difference can be captured with the simulated lung, and the movement of the lung becomes easily observable.

When visualized, the labeling method for relative scores is as follows. The image is labeled in a relative way using black and white or color matching. There is a case in which the value before and after a few percentages of density/intensity obtained by the difference is clipped, and the remaining upper and lower values are displayed relative. Alternatively, there is a case in which the value before and after a few percent of the obtained difference often becomes a significant value, and thus it is eliminated as an artifact to display in a relative way the remainder. The display is often done in the form of values from 0 to 100%, in addition to the way from 0 to 255 gradations, etc.

In addition, the pixel is rendered indeterminately to some extent, and a dark state is generated and fully displayed. Namely, in the case of the blood vessels of the lung, low signal values coexist with high signal values, but if only high signal values can be roughly obtained, it is not necessary to be generally undefined. For example, in the case of blood flow, a signal having a threshold or more may have been extracted, but in the case of respiration, a signal having a threshold or more may not have been extracted. Namely, in the case where the numerical value in the following table is taken as one pixel and the numerical value at the center is obtained, when the ratio occupied by the numerical value at the center is obtained and averaged at one pixel, its expression can be smoothly done between neighboring pixels.

1

1

1

This method is also applicable in detecting not only the lung field, but also the density in an arbitrary analysis range, and eliminating the place where the density is relatively and greatly changed. In addition, points that greatly exceed a predetermined threshold are cut off. In addition, by recognizing the shape of the ribs, for example, the sudden appearance of a high/low signal line is recognized and eliminated. In addition, there is a case where a sudden-appearing signal is similarly removed from the phase, for example, a sudden-appearing signal different from the change in the ordinary wave, which shows the sign of a patient in which an artifact is recognized in the reconstruction phase of about 15 to 20%, etc. . In addition, if we take the base data first, there is a case of another phase in the calculation of (diaphragm)^(thorax)"(chest movement)"(spirometer)"(lung field), (density) field" (volumetry), etc. d.; and there is a case where the phase is fitted to an actually recognizable shape (XP contour).

After the simulated lung is prepared as described above, it becomes possible to digitize and present the tuneability, the concordance score, and the nonconcordance score (tunable frequency image display or adjustable wavelength image). In this way, a deviation from the normal state can be displayed. According to the present embodiment, by performing a Fourier analysis, it becomes possible to determine the possibility of a new disease, comparison with oneself in a normal state, comparison of an arm with a leg, and comparison with another arm and another leg on the opposite side. In addition, it becomes possible to determine which part is abnormal in leg movement, swallowing, and the like. by digitizing customizability. In addition, it becomes possible to determine whether the person in the disease state has changed after a fixed time has elapsed, and to compare the states before and after the change in the case where the person has changed. In addition, the assessment of the inner layer - the intermediate layer, the outer layer, etc. can be easily performed, and it can be expressed as the predominance of the periphery or the predominance of the intermediate layer, making the lung field in such a shape (from round to oval) that is easily visible radially , while the distance from the periphery remains fixed.

In addition, during rendering, it is also possible to switch and display the image after the Fourier transform and the image before the Fourier transform, and both are placed and displayed on the same screen.

As shown in FIG. 2D, it becomes possible to determine what percentage difference appears in the human body when the simulated lung is set to 100 to display the change rate. In addition, it is also possible to perceive a difference not only for everything easy, but

- 18 040692 also parts of the lung. Namely, as described above, it is possible to display only the movement of the diaphragm; display the movement of the diaphragm by fixing the shape of the lung field, which differs from the diaphragm; and display a customizable concordance score and change rate. In addition, it is also possible to determine the standard blood flow by performing the Classification of Variations. Those. it becomes possible to determine the cycle of the breath element, calculate the relative relationship of the locations of the blood vessels, and determine the dynamic state of the patient's blood flow as the standard blood flow.

In addition, the lung can be detected using a pattern matching method. Each of the FIGS. 2E-2H is a diagram showing an example image of a lung field region sample. As shown in FIG. 2E-2H, the lung shape is pre-classified by pattern, and an image closely matching one of them can be extracted. This method can determine whether the target image shows one lung or both lungs. In addition, it can also be determined whether it is the left lung or the right lung. The number of samples is not limited, but it is estimated to be 4-5 samples. In addition, in this way there is also a way of recognizing the right lung, the left lung or both lungs only in the shape of the lung field. In addition, it is also possible to use the method of recognizing the right or left lung or both lungs by the relative position with the zone of decreased permeability in the form of a strip and the relative position with the zone of increased permeability of the pulmonary field by recognizing a thick zone of decreased permeability in the form of a strip along the center/mediastinum. In addition, as shown in FIG. 2H, this method can also be applied to the area on the underside of the diaphragm. In this way, the area on the underside of the diaphragm as well as the heart can also be recognized.

In addition, air has the highest permeability and is an area showing higher permeability than the lung field, and thus it is desirable to make calculations while also taking air into account. Those. the following determination can be made from the location of the air on the screen. Case (area of air at the top right of the screen)>(area of air at the top left of the screen) means the left lung. This is due to the fact that for the periphery of the shoulder, the area of air outside the human body becomes larger when photographed. Case (area of air at the top left of the screen)>(area of air at the top right of the screen) means the right lung. Similar to above, this is also due to the fact that for the periphery of the shoulder, the area of air outside the human body becomes larger when photographed. Further, case (area of air at the top right of the screen)®(area of air at the top left of the screen) means both lungs. This is due to the fact that the right and left air areas are the same size.

In addition, air from the intestinal tract often enters under the diaphragm, in which case it is possible that recognition will fail. Accordingly, it is also possible to approximately recognize first the lung field and the zone of reduced permeability on its periphery, on the sides of the mediastinum, heart, diaphragm, etc., from the central part of the lung field; and recognize the edge of the lung field or its line. Regarding this method, the technique disclosed in <https://jp.mathworks.com/help/images/examples/block-processing-large-images_ja_JP.html·>, for example, can also be used.

This allows you to compare the patient with another patient, as well as to carry out digitization. It also allows comparison of normal lung or normal blood vessels with typically abnormal lung function or abnormal blood flow and digitization. In addition, as a relative measure of lung function and lung blood flow at different times, it becomes possible for a patient to use a simulated lung and standard blood flow. The simulated lung and standard blood flow can thus be used as indexes during evaluation, which is done by morphological application to some patient by collecting representative examples of different types of typical patients and healthy individuals for simulated lung and standard blood flow.

Building an image of the lung field.

Typically, the lung field includes ribs showing low permeability and thus it is difficult to mechanically identify the contour of the lung by having density alone as an index. The present patent specification now uses a lung contour tuning method to obtain a well-matched property by pre-imaging the lung field contour using a combination of Bezier curves and straight lines.

For example, when the contour of the left lung is expressed with 4 Bezier curves and one straight line, it becomes possible to construct an image of the contour of the lung by detecting 5 points on the lung contour and 4 control points. It becomes possible to detect the contour of the lung with high accuracy by shifting the location of the point for imaging a plurality of contours of the lung and evaluating the matching property using the condition under which the total value of the density inside the contour becomes the maximum, the condition under which the difference of the sum of the density for several pixels inside and outside the line contour becomes maximum, etc. In practice, it is also possible to identify the locations of several points from the contour above the lung, the edge of which is easy to detect, and the location of the diaphragm, which is detected using the following

- 19 040692 way, thus it is possible to reduce the number of simulation attempts described above. It is also possible to extract a point near the outer edge by extracting the contour through classical binarization and adjust the position of the control point of the Bezier curve using the least squares method, etc.

Each of the FIGS. 3A and 3B is a diagram showing an example of lung field contour imaging using both Bezier curves and straight lines. Fig. 3A shows the case where the area of the lung becomes maximal (maximum contour). Fig. 3B shows the case where the lung area becomes minimal (minimum outline). In each figure, cp1~cp5 represents a control point, and p1~p5 shows a point on a Bezier curve or straight line. Thus, it becomes possible to find a contour on the path by calculating by determining the maximum contour and the minimum contour. For example, it becomes possible to display states of 10, 20% ... inhalation-exit. Thus, according to the present embodiment, it becomes possible to image at least a lung field, blood vessels, or a heart using at least one Bezier curve. In addition, the method as described above is not limited to the lungs, but is also applicable to other organs such as organ detection. In addition, for example, it is possible to perform processing to detect a range corresponding to an analysis range in another frame using at least one Bezier curve on a predetermined analysis range (tumor, brain hypothalamus, basal ganglion, inclusion boundary, etc.) in specific frame.

In addition, not only flat images but also stereoscopic images (3D images) are also applicable. It becomes possible to set a range surrounded by a plurality of curved surfaces for an analysis target by specifying an equation for a curved surface to define its reference point.

Motion detection of the dynamic zone associated with the diaphragm or breathing.

It is possible to detect the movement of the dynamic zone associated with the diaphragm or respiration in sequentially photographed images. For sequentially photographed images, when selecting images at random intervals and calculating the difference between images, the difference becomes large in the case of an area where the contrast is especially large. Areas where there has been movement can be detected by appropriate visualization of this difference. During rendering, the continuity of regions where the absolute value of the difference is large can also be emphasized by curve fitting and the like, for which denoising by thresholding, least squares, and the like is used.

According to the lung field, the contrast in the line with which the diaphragm and the heart are in contact is significant and, as shown in FIG. 4A, when the difference between two images of an easy-to-visualize difference is taken by setting a fixed threshold, a line contacting the diaphragm and the heart can be visualized, as shown in FIG. 4B.

Assessment of diaphragm movement.

The present method can detect the position of the diaphragm when the diaphragm is in motion, but it is difficult to detect the position where the motion of the diaphragm becomes weak. Those. it is difficult to detect the moments when exhalation and inhalation change places, while holding the breath, immediately after the start of photographing and immediately before its completion. In accordance with the present method, diaphragm movement is evaluated using an optional additional method.

After visualizing the line of the diaphragm, as shown in Fig. 4B, by using the above method, an image having 1024 vertical pixels was divided into 128 rectangles for every 8 vertical pixels, and the signal value contained in each rectangular area was summed to create a histogram as shown in FIG. 4C. It is assumed that the peak closest to the lower coordinate among the plurality of peaks, which is shown in the dotted box, indicates the location of the diaphragm. In keeping with the traditional XP still image, the iris is displayed as a curve, but this coordinate approximates the location of the iris.

Peak locations were detected as described in FIG. 5 by detecting the aperture location for the entire image using the present method. This detected value is corrected to estimate iris movement. First, when the difference is greater than the fixed value, it is excluded because it is treated as an outlier value (thin solid line in Fig. 5). The data, from which the outlier values were excluded, were divided into arbitrary groups, and each of the groups was subjected to bi-square curvilinear regression to connect the results to each other (solid line in Fig. 5). In the present analysis, regression analysis has been performed, but the present invention is not limited to this, and an optional additional method such as spline interpolation and the like can be used.

Refine dynamic zone detection.

There are cases where the contrast in the dynamic zone is not uniform along the line. In this case, the shape of the dynamic area can be detected more accurately by changing the threshold used to eliminate noise and performing detection processing several times. For example, for the left lung, the contrast at the diaphragm line tends to be weaker towards the inside of the human body. In FIG. 4B, only the right half of the diaphragm can be detected. In this case, the remainder, i.e. the left half of the aperture can also be detected by changing the threshold settings used to eliminate noise. It becomes possible to detect the shape of the entire diaphragm by repeating this processing several times. The present method also makes it possible to digitize not only the location of the diaphragm, but also the rate of change and the amount of change of line and surface with respect to shape, and use the new diagnostics.

It becomes possible to use the location or shape of the diaphragm thus detected. Those. in accordance with the invention of the present application, it is possible to plot the aperture coordinates; calculate the coordinates of the chest, as well as the diaphragm, using curves (surfaces) or lines calculated as described above; and depict as a graph heart contractions, blood vessel beats, density in the lung field, etc. as a position corresponding to a cycle or coordinates. This method is applicable to the dynamic zone associated with breathing.

If not only the frequency (Hz) in each exhalation and inhalation changes, but also the frequency (Hz) of the dynamic zone associated with the diaphragm or breath changes, this method allows measurements to be taken in a frequency band that responds to this change. Then, during the extraction of the BPF spectrum (bandpass filter) in a fixed range, it becomes possible to set the BBF according to each respective breathing state; it becomes possible that the optimal state can be caused by axis variation at the location of the BPF in each phase of breathing reconstruction; and it becomes possible to prepare the BPF with the variance accompanying the above. Although the breathing rhythm varies when breathing slowly or holding the breath (Hz=0), this allows images to be produced as described above.

In addition, the frequency of the entire exhalation or inhalation can be calculated based on the ratio of the breath element to the total exhalation or inhalation. In addition, those for which aperture detection is performed several times and whose signal and waveform are stable can be selected. Accordingly, it becomes possible to calculate at least one breath element frequency from the position or shape of the detected diaphragm, or the location or shape of the dynamic zone associated with the breath. It becomes possible to determine the frequency of the breath element when one can determine the location or shape of the diaphragm or dynamic zone. This method allows tracking of the subsequent waveform even though part of the waveform is broken. Thus, one can follow the original breath element even though the frequency of the breath element changes along the way. In addition, heart pulsations, etc. often undergo a sudden change, but the same can also apply to the cardiac blood vessels. Next, the operation of each module according to the present embodiment will be explained.

Analysis of respiratory function.

First, the analysis of the respiratory function will be explained. Fig. 6A is a flowchart showing the overall flow of a respiratory function analysis according to the present embodiment. The main module 1 retrieves the DICOM images from the database 15 (step S1). Here, at least a plurality of images of frames included in one breath cycle are obtained. Then, in each frame image acquired, the cycle of the breath element is determined using the density (density/intensity) in some fixed area within at least the lung field (step S2). In addition, the determined breath cycle and the waveform determined from this breath cycle can be used in each step described below.

It is also possible to further use the movement of the diaphragm and the movement of the chest to determine the cycle of the breath element. In addition, a range composed of some fixed volume and density/intensity measured in a zone showing high X-ray transmissibility and data obtained by another measurement method such as a spirogram or the like can be used. In addition, the frequency inherent in each organ (the lung in this case) is determined in advance, and the density/intensity corresponding to the determined frequency can be extracted.

Then, in FIG. 6A, the lung field is automatically detected (step S3). The contour of the lung is continuously changing, and thus, if the maximum shape and the minimum shape can be detected, it is possible to interpolate the shape between them by calculation. The lung contour in each frame image is determined by interpolating each frame image based on the cycle of the breath element determined in step S2. The lung field can be detected by performing pattern recognition as shown in FIG. 2E-2H. In addition, denoising by clipping can be performed on the detected lung field. The detected lung field is then divided into a plurality of block sections (step S4). Then, the change in each block area in each frame image is calculated (step S5). Here, the change values within each block-section are averaged and expressed as a single piece of data.

In addition, denoising by clipping can be performed for the change values within each block-section. Then, a Fourier analysis or an adjustable concordance score analysis is performed based on the cycle of the above-described breath element for the density/intensity value in each block area and the amount of change thereof (step S6).

Then denoising is performed for the results obtained by Fourier analysis or analysis of the adjusted concordance index (step S7). Here, artifact removal by clipping can be performed as described above. The above-described operations from step S5 to step S7 are performed at least once, and it is determined whether or not to complete the processing (step S8). Here, there are some cases where the frequency-tunable image of a high-purity element caused by the mixing of a compound wave and other waves, such as a breath element, a blood flow element, or another element, cannot be displayed with a single spectrum extraction for the feature amount displayed on the display. Here, there is a case where all or part of the analysis before displaying the feature value as a pixel value is repeated again or several times. With this operation, it becomes possible to additionally obtain an image of high purity in terms of adjustability and concordance of elements, such as the breath element and the blood flow element. This operation can be performed manually by the operator, visually confirming the image on the display, and it can also be performed automatically to recalculate the distribution factor by extracting the spectrum from the output. In addition, even after this calculation, noise clipping processing, void filling by least squares (interpolation), and correction using periphery density can be performed according to the situation.

At step S8, if the processing is not completed, a transition is made to step S5; and if the processing is completed, the results obtained by the Fourier analysis or the analysis with adjustable concordance index are displayed as a pseudo-color image (step S9). In addition, a black and white image may be displayed. There are some cases where the accuracy of the data is improved by repeating many cycles in this way. Thus, a desired dynamic image can be displayed. In addition, the desired dynamic image can be obtained by adjusting the image displayed on the display.

According to the present embodiment, the desired frequency or bandwidth is calculated by this calculation, but when viewing the actual image, a good image is not always displayed.

Accordingly, the following methods are often used.

(1) A method in which multiple frequency bands are presented and manually selected.

(2) A method in which multiple frequency bands are presented to extract a good image through pattern recognition with AI-based techniques.

(3) A choice is made according to the trend and shape of the HISTGRAM. Those. the value in the center of the Histgram according to the result signal tends to increase, and the value of the histgram varies according to the movement, and thus selection can be made on the trend and shape of the HISTGRAM.

Analysis of the blood flow of the lung.

Next, the lung blood flow analysis will be explained. Fig. 7 is a flowchart showing the general scheme of lung blood flow analysis according to the present embodiment. The main module 1 retrieves the DICOM images from the database 15 (step T1). Here, at least a plurality of images of frames included in one cycle of cardiac contraction are obtained. Then, the beat cycle of the blood vessels is determined based on each acquired frame image (step T2). In addition, it is possible to use the determined blood vessel beat cycle and the waveform determined from this blood vessel beat cycle in each step described below. For the beat cycle of the blood vessels, as described above, the beat of the blood vessels is analyzed using, for example, the measurement results of the electrocardiogram and other modalities, such as a tonometer, etc., and the change in the density/intensity of an arbitrary zone, such as the heart/hilum of the lung. /major blood vessels, etc. In addition, the frequency inherent in each organ (lung blood flow in this case) is determined in advance, and the density/intensity corresponding to the determined frequency can be extracted.

Next, in FIG. 7, the breath element cycle is determined in the above manner (step T3), and the lung field is automatically detected using the breath element cycle (step T4). In automatic lung contour detection, there are often variations for each frame image, but the lung contour is determined in each frame image by interpolating each frame image based on the breath element cycle determined in step T3. In addition, the lung field can be detected by performing pattern recognition as shown in FIG. 2E-2H. In addition, denoising can be performed for the detected lung field by clipping. Then the detected lung field is divided into a plurality of block-sections (stage T5). Then, the change in each block-section in each frame image is calculated (step T6). Here the meaning of the change in

- 22 040692 each block-section is averaged and expressed as one piece of data. In addition, denoising can be performed for the change value in each block-section by clipping. Then, a Fourier analysis or a variable concordance analysis is performed for the density/intensity value in each block area or the amount of change thereof based on the beat cycle of the blood vessels described above (step T7).

Denoising is then performed for the results obtained using Fourier analysis or analysis with an adjustable concordance index (step T8). Here, artifact removal by clipping can be performed as described above. The above-described operations from step T6 to step T8 are performed at least once, and it is determined whether or not to end the processing (step T9). Here, there are some cases where the frequency-tunable image of a high-purity element caused by the mixing of a compound wave and other waves, such as a breath element, a blood flow element, or another element, cannot be displayed with a single spectrum extraction for the feature amount displayed on the display. Here, there is a case where all or part of the analysis before displaying the feature value as a pixel value is repeated again or several times. With this operation, it becomes possible to additionally obtain an image of high purity in terms of adjustability and concordance of elements, such as the breath element and the blood flow element. This operation can be performed manually by the operator, visually confirming the image on the display, and it can also be performed automatically to recalculate the distribution factor by extracting the spectrum from the output. In addition, even after this calculation, noise clipping processing, void filling by least squares (interpolation), and correction using periphery density can be performed according to the situation.

At step T9, if the processing is not completed, a transition is made to step T6; and if the processing is completed, the results obtained by Fourier analysis or analysis with adjustable concordance index are displayed as a pseudo-color image (step T10). In addition, a black and white image may be displayed. Thus, it becomes possible to improve the accuracy of the data. In addition, the desired dynamic image can be obtained by adjusting the image displayed on the display.

According to the present embodiment, the required frequency or bandwidth is calculated by this calculation, but when viewing the actual image, a good image is not always displayed. Accordingly, the following methods are often used.

(1) A method in which multiple frequency bands are presented and manually selected.

(2) A method in which multiple frequency bands are presented to extract a good image through pattern recognition with AI-based techniques.

(3) A choice is made according to the trend and shape of the HISTGRAM. Those. the value at the center of the Histgram according to the result signal tends to increase, and the value of the histgram varies according to the movement, and thus selection can be made on the trend and shape of the HISTGRAM.

Analysis of other blood flow.

Next, the analysis of another blood flow will be explained. As shown in FIG. 15, one aspect according to the present invention can also be applied to the analysis of the blood flow of the heart, aorta, lung blood vessels, brachial artery, cervical blood vessels, and the like. In addition, blood flow analysis is also similarly applicable to abdominal blood vessels, peripheral blood vessels, etc., which are not shown in the diagram. Fig. 8 is a flowchart showing the general scheme for analyzing other blood flow according to the present embodiment. The main module 1 retrieves the DICOM images from the database 15 (step R1). Here, at least a plurality of images of frames included in one cycle of cardiac contraction are obtained. Then, the beat cycle of the blood vessels is determined based on each acquired frame image (step R2). In addition, the determined blood vessel beat cycle and the waveform determined from this blood vessel beat cycle can be used in each step described below. For the beat cycle of the blood vessels, as described above, the beat of the blood vessels is analyzed using, for example, the measurement results of the electrocardiogram and other modalities, such as a tonometer, etc., and the change in the density/intensity of an arbitrary zone, such as the heart/hilum of the lung. /major blood vessels, etc. In addition, the frequency inherent in each organ (eg, major blood vessels) is determined in advance, and the density/intensity corresponding to the detected frequency can be extracted.

Then, an analysis range is set (step R3), and the set analysis range is divided into a plurality of block sections (step R4). The value of the change in each block-section is then averaged and expressed as a single piece of data. In addition, denoising can be performed for the change value in each block-section by clipping. Fourier analysis or adjustable concordance analysis is then performed based on the beat cycle of blood vessels described above for the density/intensity value in each block area and the amount of change thereof (step R5).

Then denoising is performed for the results obtained using Fourier analysis or analysis with a customizable concordance index (step R6). Artifact removal by clipping can be performed here, as described above. The above-described operations from step R5 to step R6 are performed at least once, and it is determined whether or not to end the processing (step R7). Here, there are some cases where the frequency-tunable image of a high-purity element caused by the mixing of a composite wave and other waves, such as a breath element, a blood flow element, or another element, cannot be displayed with a single spectrum extraction for the feature amount displayed on the display. Here, there is a case where all or part of the analysis before displaying the feature value as a pixel value is repeated again or several times. With this operation, it becomes possible to additionally obtain an image of high purity in terms of adjustability and concordance of elements, such as the breath element and the blood flow element. This operation can be performed manually by the operator, visually confirming the image on the display, and it can also be performed automatically to recalculate the distribution factor by extracting the spectrum from the output. In addition, even after this calculation, noise clipping processing, void filling by least squares (interpolation), and correction using periphery density can be performed according to the situation.

At step R7, if the processing is not completed, a transition is made to step R5; and if the processing is completed, the results obtained by Fourier analysis or analysis with adjustable concordance index are displayed as a pseudo-color image (step R8). In addition, a black and white image may be displayed. Thus, it becomes possible to improve the accuracy of the data. In addition, the desired dynamic image can be obtained by adjusting the image displayed on the display.

According to the present embodiment, the required frequency or bandwidth is calculated by this calculation, but when viewing the actual image, a good image is not always displayed. Accordingly, the following methods are often used.

(1) A method in which multiple frequency bands are presented and manually selected.

(2) A method in which multiple frequency bands are presented to extract a good image through pattern recognition with AI-based techniques.

(3) A choice is made according to the trend and shape of the HISTGRAM. Those. the value at the center of the Histgram according to the result signal tends to increase, and the value of the histgram varies according to the movement, and thus selection can be made on the trend and shape of the HISTGRAM.

In addition, if the analysis is performed in 3D, it becomes possible to calculate the inspiratory output volume, cardiac output, and central blood flow in each block area from the Fourier analysis result as a relative value by measuring the inspiratory-expiratory volume, cardiac output, and central blood flow using another devices. Those. in the case of analysis of respiratory function, it becomes possible to assess the volume of ventilation of the lung by the volume of inhalation-exhalation; in the case of analysis of the blood flow of the lung, it becomes possible to estimate the amount of blood flow to the lung by cardiac (blood vessels of the lung) ejection; and in the case of analyzing the amount of other blood flow, it becomes possible to estimate the estimated amount of blood flow (indicator) in branching blood vessels, which is taken from the amount of blood flow (indicator) on the central side.

In addition, as described above, the determination can be made with higher accuracy if the entire resulting database can be calculated, but this often takes time when performing computer analysis. Thus, it is possible to perform calculations by extracting only an arbitrary number of them (for example, a specific phase). By this, the analysis time can be reduced, and, in addition, irregularities observed at the beginning of the breath can be cut out. In addition, an arbitrary range can be displayed when displaying the analysis result. For example, the so-called infinite playback is provided during the repeated playback, and it becomes possible for the doctor to easily perform diagnosis by displaying the expiration/inspiration change range from the exhalation/inspiration change.

As described above, according to the present embodiment, it becomes possible to evaluate human body images using an X-ray dynamic imaging apparatus. If digital data can be obtained, the calculations can generally be performed well with existing facility devices, and thus the installation cost is reduced. For example, for a dynamic X-ray imaging apparatus with a flat panel detector, it becomes possible to easily perform an examination of a patient. In addition, fluoroscopy of an embolism or thrombosis of the lung for lung blood flow becomes possible. For example, for a dynamic x-ray imaging device provided with a flat panel detector, redundant examinations can be eliminated by executing a diagnostic support program in accordance with

Claims (34)

worthwhile embodiment prior to performing CT. In addition, a simple examination is performed, and thus it becomes possible to detect a disease with a high degree of urgency at an early stage and treat it preferentially. According to the way of photography at present, in case of other modality such as CT, MR, etc., there are some problems, but a detailed diagnosis in each area is possible if the above problems can be solved. In addition, it is also applicable to fluoroscopy of various kinds of blood vessels, such as cervical constriction of blood flow; and also applicable to blood flow assessment and fluoroscopy of large blood vessels. In addition, with regard to respiration data, they allow partial lung function examinations and can also be used for lung function examinations. In addition, it also becomes possible to identify diseases such as chronic obstructive pulmonary disease (COPD), emphysema, and the like. It can also be used to determine characteristic conditions before and after surgery. In addition, by Fourier transforming the cycle of the respiration element and the blood flow cycle and eliminating the respiration waveform and the blood flow waveform in the abdominal X-ray, it becomes possible to detect pathologies in the remaining biological motion such as intestinal obstruction and the like. In addition, when an image obtained at the beginning exhibits a high resolution to some extent, there are a large number of pixels, and thus the calculation often takes time. In this case, the calculation can be performed after reducing the image to a fixed number of pixels. For example, the calculation time can be reduced by performing the calculation after reducing the number of pixels from [4096x4096] to [1024x1024] pixels. Other. In addition, when photographing to obtain an X-ray, for example, a prediction algorithm such as an AR method (Autoregressive Moving Average Model), etc., may be used. When at least one frequency of the breath element can be identified, it is also possible to control the X-ray photographing device to adjust the X-ray irradiation interval according to that frequency. For example, in the case of a low frequency of the breath element (in the case of a long cycle), the number of X-ray photographs can be reduced. Through this, it becomes possible to reduce the amount of radiation exposure of the human body. In addition, in the case of a high frequency of the respiratory element or the cardiovascular element, such as tachypnea, tachycardia, and the like. (in the case of a short cycle), an optimal image can be obtained by increasing the exposure frequency. In addition, according to the DICOM data storage format, the quality of a compressed image is often degraded, and thus it is preferable to store it without being compressed. In addition, the calculation method can be changed according to the data compression format. Explanation of symbols 1 - Main module; 3 - block analysis of respiratory function; 5 - block analysis of blood flow of the lung; 7 - block analysis of another blood flow; 9 - Fourier analysis block; 10 - waveform analysis unit; 11 - visualization / digitization block; 13 - input interface; 15 - database; 17 - output interface; 19 - display. CLAIM 1. A system for diagnosing the state of breathing and the state of the blood vessels of the human body, analyzing images of the human body and displaying the results of the analysis, and the system contains an input interface configured to receive images of human body frames from an x-ray dynamic imaging device; the main module, configured to detect the organ field in each frame image, divide the organ field image in each frame image into block-sections of the organ field image and obtain the calculated change in the image in each block-section in each frame image according to the cyclic change characteristic of this organ in the result of breathing in the form of a breathing waveform; for each block-section of the image of the field of the organ in each image of the frame, perform the Fourier transform for the calculated image change, obtain the extracted frequencies close to the cyclic change characteristic of this organ as a result of breathing, and perform the inverse - 25 040692 Fourier transform for the extracted frequencies to obtain a change in the image tuned to respiration, detect the field of the mentioned or other organ in each image of the frame, split the image of blood vessels in the field of the mentioned or other organ in each image of the frame into blocks of the image of blood vessels and obtain the calculated change of the image in each block-section in each image of the frame according to the cyclic change inherent in the blood vessels of the mentioned or another organ as a result of heart contractions or beating of blood vessels in the form of a waveform of heart contractions or a waveform of the beating of blood vessels, for each block-section of the image of blood vessels in the field of the mentioned or other organ in each image of the frame, perform the Fourier transform for the calculated image change, obtain the extracted frequencies close to the cyclic characteristic of the blood vessels of the mentioned or other organ, and change due to heartbeat or blood vessel beat, and perform an inverse Fourier transform on these extracted frequencies to obtain an image change tuned to heart beat or blood vessel beat; perform digitization using Fourier analysis data to obtain processed images representing the actual movement; an output interface configured to output at least one processed image to the display, wherein the main module is additionally configured to detect the frequency of changes in the image of an organ or the frequency of changes in the image of blood vessels of the said or another organ from the received frame images in each block-section. 2. The system of claim 1, wherein an image showing organ cycling is output, or images other than the image showing organ cycling are output. 3. The system of claim 1, wherein the filter is a filter configured to extract, in each block-section, a frequency corresponding to a heart image density change, a lung image density change, or a lung or heart blood vessel image density change. 4. A system for diagnosing the state of breathing and the state of the blood vessels of the human body, analyzing images of the human body and displaying the results of the analysis, and the system contains an input interface configured to receive images of frames of the human body from an X-ray dynamic imaging device, and the frame images are obtained for tracking an arbitrary section of the organ according to frame images; and the main module, configured to detect the organ field in each frame image, divide the organ field image in each frame image into block areas of the organ field image, and obtain the calculated change in image density in each block area in each frame image according to the cyclic characteristic of this organ change as a result of breathing in the form of a breathing waveform; for each block-section of the organ field image in each frame image, perform a Fourier transform for the calculated change in image density, obtain extracted frequencies close to the cyclic change due to respiration inherent in this organ, and perform an inverse Fourier transform for the extracted frequencies to obtain an image change tuned for breathing; detect the organ field in each frame image, divide the blood vessel image in the organ field in each frame image into block-sections of the blood vessel image, and obtain the calculated change in image density in each block-section in each frame image according to the cyclic change inherent in the blood vessels of the organ as a result heart beats or blood vessel beats in the form of a heart beat waveform or a blood vessel beat waveform; for each block-section of the image of blood vessels in the field of the organ in each image of the frame, perform the Fourier transform for the calculated change in image density, obtain the extracted frequencies close to the cyclic change inherent in the blood vessels of the organ as a result of heart contractions or the beating of blood vessels, and perform the inverse Fourier transform for these extracted frequencies to obtain an image change tuned to heart beats or blood vessel beats; perform digitization using Fourier analysis data to obtain processed images representing the actual movement; output interface configured to output at least one processed image to the display, wherein the main module is additionally configured to monitor each of the said block-sections of the organ image and the said blocks-sections of the image of the blood vessels of the organ in a state of maintaining relative mutual the location of each of the block sections between frame images. 5. A system for diagnosing the state of respiration and the state of the blood vessels of the human body, analyzing the images of the human body and displaying the results of the analysis, and the system contains an input interface configured to receive images of human body frames from an X-ray dynamic imaging device; the main module, configured to detect the lung field in each frame image, divide the lung field in each frame image into block areas of the lung field image, and obtain the calculated change in the image in each block area in each frame image according to the inherent slight cyclic change as a result of breathing in the form of a breath waveform; for each lung field image block in each frame image, perform a Fourier transform for the computed image change, select at least one extracted frequency corresponding to respiration, and perform an inverse Fourier transform for said at least one extracted frequency corresponding to respiration, wherein said at least one extracted frequency is selected based on the spectral structural relationship in the light cycling characteristic of the spectra obtained after the Fourier transform; detect the lung field in each frame image, divide the blood vessel image in the lung field in each frame image into blood vessel image blocks, and obtain the calculated image change in each block region in each frame image according to the light cycling characteristic of blood vessels due to cardiac contractions or beats of blood vessels in the form of a waveform of heart contractions or a waveform of the beat of blood vessels; for each block-section of the image of blood vessels in the pulmonary field in each frame image, perform a Fourier transform for the calculated image change, select at least one extracted frequency corresponding to heart contractions or blood vessel beats, and perform an inverse Fourier transform for the at least one mentioned a extracted frequency corresponding to a heartbeat or a beating of a blood vessel, wherein said at least one extracted frequency is selected based on a spectral structural relationship in blood vessel light cycling from spectra obtained after a Fourier transform; perform digitization using Fourier analysis data to obtain processed images representing the actual movement; an output interface configured to output at least one processed image to a display. 6. A system for diagnosing the state of respiration and the state of the blood vessels of the human body, analyzing images of the human body and displaying the results of the analysis, and the system contains an input interface configured to receive images of frames of the human body from an X-ray dynamic imaging device; the main module, configured to detect the lung field in each frame image, divide the lung field in each frame image into block areas of the lung field image, and obtain the calculated change in the image in each block area in each frame image according to the inherent slight cyclic change as a result of breathing as a waveform of the breath element, wherein the breath element includes all or part of a person's exhalation or inhalation; for each block-section of the lung field image in each image of the frame, perform the Fourier transform for the calculated image change for the breath element, obtain at least one extracted frequency of the breath element, and the said at least one extracted frequency of the breath element is extracted in the band corresponding to the breath, perform an inverse Fourier transform on said at least one extracted breath element frequency; detect the heart field in each frame image, split the blood vessel image in the heart field in each frame image into heart blood vessel image blocks, and obtain the computed image change in each block region in each frame image from the heart-specific cyclic change due to heart contraction or blood vessel beats in the form of a waveform of a cardiovascular element, wherein the cardiovascular element is extracted from heart beats or blood vessel beats of the human body; for each block-section of the image of blood vessels in the field of the heart in each image - 27 040692 frames to perform a Fourier transform for the calculated image change for the cardiovascular element, to obtain at least one extracted frequency of the cardiovascular element, wherein said at least one extracted frequency of the cardiovascular element is extracted in a band corresponding to heart contractions or beats blood vessels, performing an inverse Fourier transform on said at least one extracted cardiovascular element frequency; perform digitization using Fourier analysis data to obtain processed images representing the actual movement; an output interface configured to output at least one processed image to the display, wherein the band corresponding to respiration is in the range from 0 to 0.5 Hz, and the band corresponding to heart contractions or beating of blood vessels is in the range from 0.6 up to 2.5 Hz. 7. The system according to claim 6, wherein the main module is additionally configured to extract spectra in a fixed band, including the noise frequency and including spectra corresponding to a frequency different from the frequency of the respiratory element or the cardiovascular element obtained from frame images , or the input frequency or bandwidth, from the spectra obtained after the Fourier transform, by using a filter. 8. The system according to claim 6 or 7, wherein the main module is further configured to generate an image between frames based on the frequency of the respiratory or cardiovascular element and each of the images of the frames. 9. A system for diagnosing the state of breathing and the state of the blood vessels of the human body, analyzing images of the human body and displaying the results of the analysis, and the system contains an input interface configured to receive images of frames of the human body from an X-ray dynamic imaging device; the main module, configured to detect the lung field in each frame image, divide the lung field in each frame image into block areas of the lung field image, and obtain the calculated change in image density in each block area in each frame image according to the inherent slight cyclic change as a result breathing; as a waveform of the breath element, wherein the breath element includes all or part of a person's exhalation or inhalation; for each block-section of the lung field image in each frame image, perform the Fourier transform for the calculated change in image density for the respiratory element, extract the spectra in the band, which includes the spectra corresponding to at least one frequency of the respiratory element, from the spectra obtained after the Fourier transform , and perform the inverse Fourier transform for the spectra extracted from this band; detect the organ field in each frame image, divide the image of blood vessels in the organ field in each frame image into block-sections of the image of the blood vessels of the organ and obtain the calculated change in image density in each block-section in each frame image according to the cyclic change inherent in the blood vessels of the organ the result of contraction of the heart or beating of blood vessels in the form of a waveform of the cardiovascular element, and the cardiovascular element is extracted from the heart contractions or beating of blood vessels of the human body; for each block-section of the image of blood vessels in the field of the organ in each image of the frame, perform a Fourier transform for the calculated change in image density for a cardiovascular element, extract spectra in a band that includes spectra corresponding to at least one frequency of a cardiovascular element, from the spectra obtained after the direct Fourier transform, and perform the inverse Fourier transform for the spectra extracted from this band; perform digitization using Fourier analysis data to obtain processed images representing the actual movement; an output interface configured to output at least one processed image to a display. 10. A system for diagnosing the state of breathing and the state of the blood vessels of the human body, analyzing images of the human body and displaying the results of the analysis, and the system contains an input interface configured to receive images of frames of the human body from an X-ray dynamic imaging device; the main module, configured to detect the lung field in each frame image by matching with the sample, divide the lung field in each frame image into block areas of the lung field image to average the pixel value of the image change, and obtain the calculated image change in each block area in each frame image according to the inherent light cyclic - 28 040692 change as a result of breathing in the form of a waveform of the breath element, and the breath element includes all or part of the exhalation or inhalation of a person; for each block-section of the lung field image in each image of the frame, perform the Fourier transform for the calculated image change for the respiratory element, extract the spectra in the first band, which includes the spectra corresponding to at least one frequency of the respiratory element, from the spectra obtained after the Fourier transform , and perform the inverse Fourier transform for the spectra extracted from the first band; detect the pulmonary field in each frame image by pattern matching, split the image of the blood vessels in the pulmonary field in each frame image into block areas of the image of blood vessels in the pulmonary field to average the pixel value of the image change, and obtain the calculated image change in each block area in each image a frame by slight cycling inherent in contraction of the heart or beating of blood vessels in the form of a waveform of the cardiovascular element, and the cardiovascular element is extracted from the heartbeats or beating of the blood vessels of the human body; for each block-section of the image of blood vessels in the pulmonary field in each image of the frame, perform the Fourier transform for the calculated image change for the cardiovascular element, extract spectra in the second band, including spectra corresponding to at least one frequency of the cardiovascular element, from the spectra obtained after the forward Fourier transform, and perform the inverse Fourier transform for the spectra extracted from the second band; perform digitization using Fourier analysis data to obtain processed images representing the actual movement; output interface configured to output at least one processed image to the display, wherein the main module is additionally configured to track each of the mentioned blocks-sections of the image of the lung field and the mentioned blocks-sections of the image of blood vessels in a state of maintaining the relative relative position of each of the block sections between frame images, wherein the main module is further configured to reconstruct a normal lung or normal blood vessels and detect pathology by comparing the lung or blood vessels of a human lung with a normal lung or normal blood vessels. 11. The system according to claim 9 or 10, wherein the main module is further configured to extract spectra in a fixed band including the noise frequency and including spectra corresponding to a frequency different from the frequency of the element of cardiovascular contractions obtained from frame images , or the input frequency or bandwidth, from the spectra obtained after the Fourier transform, by using a filter. 12. The system according to any one of claims 9 to 11, wherein the main module is further configured to generate an image between frames based on the frequency of the respiratory element or the cardiovascular element and each of the frame images. 13. A system for diagnosing the state of respiration and the state of the blood vessels of the human body, analyzing images of the human body and displaying the results of the analysis, the system comprising an input interface configured to receive images of frames of the human body from an x-ray dynamic imaging device; the main module, configured to detect the lung field in each frame image by matching with the sample, divide the lung field in each frame image into block areas of the lung field image to average the pixel value of the image change, and obtain the calculated image change in each block area in each displaying the frame by the inherent slight cyclic change as a result of breathing in the form of a waveform of the breath element, and the breath element includes all or part of the exhalation or inhalation of a person; for each block-section of the lung field image in each image of the frame, perform the Fourier transform for the calculated image change for the respiratory element, extract the spectra in the first band, which includes the spectra corresponding to at least one frequency of the respiratory element, from the spectra obtained after the Fourier transform , and perform the inverse Fourier transform for the spectra extracted from the first band; detect the heart field in each frame image by pattern matching, split the image of blood vessels in the heart field in each frame image into block areas of the image of blood vessels in the heart field to average the pixel value of the image change, and obtain the calculated image change in each block area in each frame image according to the cyclic change inherent in the blood vessels of the heart as a result of the contraction of the heart or the beating of the blood vessels in the form of a waveform of the cardiovascular element, and the cardiovascular element is extracted from the heart contractions or the beating of the blood vessels of the human body; for each block-section of the image of blood vessels in the field of the heart in each image of the frame, perform the Fourier transform for the calculated image change for the cardiovascular element, extract spectra in the second band, including spectra corresponding to at least one frequency of the cardiovascular element, from the spectra obtained after the forward Fourier transform, and perform the inverse Fourier transform for the spectra extracted from the second band; perform digitization using Fourier analysis data to obtain processed images representing the actual movement; an output interface configured to output at least one processed image to the display, wherein the main module is additionally configured to track each of said lung field image blocks and said blood vessel image blocks in the heart field in a state of maintaining relative relative position each of the block-sections between frame images, wherein the main module is additionally configured to reconstruct a normal lung or normal blood vessels and detect pathology by comparing a human lung or heart blood vessels with a normal lung or normal blood vessels. 14. The system according to claim 13, wherein the main module is further configured to extract spectra in a fixed band including the noise frequency and including spectra corresponding to a frequency different from the frequency of the breath element or the blood vessel beat element obtained from frame images , or the input frequency or bandwidth, from the spectra obtained after the Fourier transform, by using a filter. 15. The system according to claim 13 or 14, wherein the main module is further configured to generate an image between frames based on the frequency of the respiratory element or the cardiovascular element and each of the images of the frames. 16. A system for diagnosing the state of breathing and the state of the blood vessels of the human body, analyzing images of the human body and displaying the results of the analysis, and the system includes an input interface configured to receive images of frames of the human body from an x-ray dynamic imaging device; the main module, configured to detect the lung field in each frame image, divide the lung field in each frame image into block areas of the lung field image to average the pixel value of the image change, and obtain the calculated image change in each block area in each frame image according to the characteristic a slight cycling change as a result of breathing in the form of a breath element waveform, wherein the breath element includes all or part of a person's exhalation or inhalation; for each block-section of the lung field image in each image of the frame, perform the Fourier transform for the calculated image change for the respiratory element, extract the spectra in the first band, which includes the spectra corresponding to at least one frequency of the respiratory element, from the spectra obtained after the Fourier transform , and perform the inverse Fourier transform for the spectra extracted from the first band; detect the lung field in each frame image, split the image of blood vessels in the lung field in each frame image into block areas of the image of blood vessels in the lung field to average the pixel value of the image change, and obtain the calculated image change in each block area in each frame image by a light cycling characteristic of blood vessels due to contraction of the heart or beating of blood vessels in the form of a waveform of a cardiovascular element, wherein the cardiovascular element is derived from heart contractions or beating of blood vessels of the human body; for each block-section of the image of blood vessels in the pulmonary field in each image of the frame, perform the Fourier transform for the calculated image change for the cardiovascular element, extract spectra in the second band, including spectra corresponding to at least one frequency of the cardiovascular element, from the spectra obtained after the forward Fourier transform, and perform the inverse Fourier transform for the spectra extracted from this band; perform digitization using Fourier analysis data to obtain processed images representing the actual movement; an output interface configured to output at least one processed image to a display, wherein the main model is further configured to calculate a pixel change rate in each block area in each frame image; extract only the block-sections for which the adjustable index is within a predetermined fixed range, and the adjustable index is the ratio of the change index of pixels in each block-section to the change index of the dynamic zone associated with breathing, and the output interface is further configured to output display the processed image, which includes only the extracted blocks-sections. 17. The system of claim 16, wherein said at least one respiratory element frequency is in the range of 0 to 0.5 Hz, wherein said at least one cardiovascular contraction element frequency is in the range of 0.6 to 2.5 Hz . 18. The system according to claim 16 or 17, wherein the logarithmic values of the adjustable indicator are defined as a fixed range including 0. 19. The system according to any one of claims 1 to 18, wherein the main module is further configured to determine the analysis range for all received frame images using a Bezier curve; and detecting an analysis target based on a change in intensity in the analysis range. 20. The system of claim 19, wherein the main module is further configured to calculate a feature of the periphery of the detected analysis target. 21. A system according to any one of claims 6 to 12, wherein internal reference points are selected in the detected lung field and the lung field is subdivided using curved or straight lines passing through the internal reference points within the lung field. 22. The system of claim 21, wherein the interval between test points in the outer extent of the detected lung field is made relatively large and the interval between internal test points is made relatively small in accordance with the expansion rate for each block-site within the detected lung field. 23. The system according to claim 21, wherein the interval between the control points is made relatively large and large as they approach in the craniocaudal direction towards the human body, or relatively large and large as they approach a particular vector direction in the detected lung field. 24. The system according to any one of claims 1 to 23, wherein the main module is further configured to plot at least one of a lung field, blood vessels, or heart using at least one Bezier curve. 25. The system of any one of claims 1 to 24, wherein the diaphragm is detected by calculating the difference in intensity for each of the continuous images, and an index is displayed indicating the location or shape of the dynamic zone associated with the detected diaphragm or breath. 26. The system of claim 25, wherein an aperture that is not interrupted by other zones other than the aperture is displayed by changing the intensity threshold to interpolate the entire shape of the aperture. 27. The system of claim 25 or 26, wherein the main module is further configured to calculate at least one breath element frequency from the location or shape of the detected diaphragm or the location or shape of the dynamic region associated with the breath. 28. The system according to any one of claims 6 to 9, wherein the breath element is corrected by changing the phase of at least one frequency of the breath element or by smoothing the waveform of the breath element. 29. The system of any one of claims 1 to 28, wherein the waveform of any section block is determined from the section blocks within the analysis range, and frequency constituents of the determined waveform are extracted to display an image corresponding to the frequency constituents of the waveform. 30. A system according to any one of claims 1 to 29, wherein density is detected in the analysis range and the location where the density changes in a relatively large manner is removed from there. 31. The system according to any one of claims 6 to 15, wherein the main module is further configured to select at least one frequency when performing an inverse Fourier transform based on a spectral structural relationship in the inherent organ cycling from the spectra obtained after the Fourier transform. 32. A system according to any one of claims 6 to 9, wherein the x-ray photographing device is controlled to adjust the x-ray irradiation interval in accordance with said at least one frequency of the breath element. 33. The system according to any one of claims 6 to 15, wherein after performing the inverse Fourier transform, only the block in which the amplitude value is relatively large is extracted and displayed. 34. The system according to any one of claims 6-12, wherein the main module is additionally configured to determine the diaphragm or chest after identification of the lung field, calculate the value -