Article Per Year (5 Year)
p-Index From 2020 - 2025
0.817
P-Index
This Author published in this journals
All Journal Journal of Electronics, Electromedical Engineering, and Medical Informatics Indonesian Journal of Electronics, Electromedical Engineering, and Medical Informatics Jurnal Teknokes
Huynh, Phuoc-Hai
Unknown Affiliation
Biochemistry, Genetics & Molecular Biology Computer Science & IT Control & Systems Engineering Decision Sciences, Operations Research & Management Electrical & Electronics Engineering Engineering Environmental Science Health Professions Materials Science & Nanotechnology

Published : 4 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 4 Documents
Search

 1 
Comparison of CatBoost and Random Forest Methods for Lung Cancer Classification using Hyperparameter Tuning Bayesian Optimization-based Zamzam, Yra Fatria; Saragih, Triando Hamonangan; Herteno, Rudy; Muliadi; Nugrahadi, Dodon Turianto; Huynh, Phuoc-Hai
Journal of Electronics, Electromedical Engineering, and Medical Informatics Vol 6 No 2 (2024): April
Publisher : Department of Electromedical Engineering, POLTEKKES KEMENKES SURABAYA

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35882/jeeemi.v6i2.382

Abstract

Lung Cancer is a disease that has a high mortality rate and is often difficult to detect until it reaches a very severe stage. Data indicates that lung cancer cases are typically diagnosed late, posing significant challenges to effective treatment. Early detection efforts offer potential for better recovery chances. Therefore, this research aims to develop methods for the identification and classification of lung cancer in the hope of providing further knowledge on effective ways to detect this condition at an early stage. One approach under scrutiny involves employing machine learning classification techniques, anticipated to serve as a pivotal tool in early disease detection and enhancing patient survival rates. This study involves five stages: data collection, data preprocessing, data partitioning for training and testing using 10-fold cross validation, model training, and analysis of evaluation results. In this research, four experiments consist of applying two classification methods, CatBoost and Random Forest, each tested using default hyperparameter and hyperparameter tuning using Bayesian Optimization. It was found that the Random Forest model using hyperparameter tuning Bayesian Optimization outperformed the other models with accuracy (0.97106), precision (0.97339), recall (0.97185), f-measure (0.97011), and AUC (0.99974) for lung cancer data. These findings highlight Bayesian Optimization for hyperparameter tuning in classification models can improve clinical prediction of lung cancer from patient medical records. The integration of Bayesian Optimization in hyperparameter tuning represents a significant step forward in refining the accuracy and effectiveness of classification models, thus contributing to the ongoing enhancement of medical diagnostics and healthcare strategies.
Privacy-Preserving Healthcare Analytics in Indonesia Using Lightweight Blockchain and Federated Learning: Current Landscape and Open Challenges Mardiansyah, Viddi; Bayuaji, Luhur; Herlistiono, Iwa Ovyawan; Violina, Sriyani; Purnama, Adi; Prasetyo, Bagus Alit; Huynh, Phuoc-Hai
Indonesian Journal of Electronics, Electromedical Engineering, and Medical Informatics Vol. 7 No. 2 (2025): May
Publisher : Jurusan Teknik Elektromedik, Politeknik Kesehatan Kemenkes Surabaya, Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35882/ijeeemi.v7i2.63

Abstract

Healthcare data are invaluable assets in today’s digital age; however, they are also highly vulnerable to misuse, breaches, and unauthorized access. The global healthcare sector faces a significant dilemma: To leverage exceptionally enormous and heterogeneous datasets, the protection of patient privacy must be ensured while simultaneously improving medical services and public health understanding. In recent years, blockchain technology has emerged as a promising solution to manage healthcare data in a decentralized, transparent, tamperproof, as well as secure way. However, several natural limitations often obstruct many conventional blockchain systems. These limitations include scalability issues, high energy consumption, in addition to increased latency, and they can greatly impede practical adoption in resource-limited settings, particularly in developing countries such as Indonesia. These many limitations considerably spurred developers to create lightweight blockchain frameworks. These frameworks aim to retain all of the core benefits of blockchain, such as its immutability in addition to traceability, and optimize both performance and efficiency. In the event that an individual integrates the proposed system by means of federated learning, which allows training of machine learning models across distributed data sources without data privacy being compromised, the system subsequently offers a compelling solution for healthcare analytics that preserves privacy in its entirety. This paper explores integrated technologies in Indonesian healthcare and highlights their potential and limitations. This study discusses how data can improve services while protecting patient confidentiality despite increasing cyber threats. It also considers regional policies like the Personal Data Protection Law and the BPJS health insurance. Identified are certain open challenges, in addition to particular future research directions, for the purpose of addressing the practical, technical, and regulatory hurdles that must be overcome to realize secure and privacy-aware healthcare analytics in Indonesia.
Design of Incu Analyzer for IoT-based Baby Incubator Calibration Septiana, Silvi Dwi; Syaifudin, Syaifudin; Maghfiroh, Anita Miftahul; Huynh, Phuoc-Hai
Jurnal Teknokes Vol. 16 No. 3 (2023): September
Publisher : Jurusan Teknik Elektromedik, Politeknik Kesehatan Kemenkes Surabaya, Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

Incubator Analyzer is a calibration tool that measures baby incubators' temperature, mattress temperature, humidity, airflow, and sound level. This research aims to design an "Incu Analyzer for IOT-based Baby Incubator Calibration (Chamber Temperature)" tool with an LCD and Thingspeak integration. The design of this calibration tool involves the baby incubator chamber temperature parameters, namely T1, T2, T3, T4, and T5, which are measured using the DS18B20 sensor. The ESP32 microcontroller is employed to leverage the IoT system, and Wi-Fi is used for IoT communication. The ESP32 processes data collected from the DS18B20 temperature sensor and displays it on the LCD and Thingspeak. This tool is tested by comparing the incubator analyzer module with a standard measuring instrument, INCU II. The temperature parameter yielded the smallest error value of -0.059293% at T5 with a setting temperature of 36°C and the largest error value of -0.0254188% at T2 with a setting temperature of 35°C. In conclusion, after conducting a comprehensive study of the literature and planning, it can be affirmed that the "Incu Analyzer Design for IOT-Based Baby Incubator Calibration" tool functions as planned, demonstrating its efficacy as an IoT-based Incubator Analyzer. This research has successfully developed an IoT system that utilizes Wi-Fi to transmit data and display reading results on Thingspeak, which significantly facilitates users in monitoring the calibration process.
Design and Development of SpO2, Bpm, and Body Temperature for Monitoring Patient Conditions in IOT-Based Special Isolation Rooms Purwitosari, Dyah; Irianto, Bambang Guruh; Triwiyanto, Triwiyanto; Huynh, Phuoc-Hai
Jurnal Teknokes Vol. 16 No. 2 (2023): June
Publisher : Jurusan Teknik Elektromedik, Politeknik Kesehatan Kemenkes Surabaya, Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

The utilization of batteries as the primary power source in portable equipment systems presents certain drawbacks, primarily concerning the need for constant monitoring of battery power to ensure uninterrupted system functionality. Therefore, this study aims to address the battery power efficiency analysis to evaluate the viability of portable systems. The research endeavors to develop a portable measurement system capable of monitoring SPO2 (blood oxygen saturation), BPM (beats per minute), and body temperature in a specialized isolation treatment room. The proposed system is designed to assess the health conditions of patients afflicted with infectious diseases by measuring their heart rate, body temperature, and oxygen saturation. The devised measurement system incorporates a 2200mAH battery to power the IC TTGO ESP32, which manages data and displays measurement results. Additionally, the system integrates the MAX30102 sensor to measure oxygen saturation and heart rate, along with the MCP9808 sensor to monitor body temperature. To ensure its accuracy, the designed device underwent rigorous testing on respondents aged 25-40 years. The sensors were placed on the fingertip, and the resulting measurements were compared against those obtained from a standardized and calibrated device. The analysis of the measurement results exhibited a commendable ±5% error margin, indicating the feasibility of the proposed device for practical usage. Moreover, the study scrutinized the efficiency of battery power utilization in two distinct modes: normal mode and save mode. In the normal mode, the device consumed a current of 154.9 mA, while the save mode, which involved deactivating the LCD TTGO ESP32, required a current of 126.7 mA. The findings demonstrated that the device could operate for approximately ±14 hours in normal mode and up to ±17 hours in save mode before the battery needed recharging. The proposed design presents an effective approach for evaluating power efficiency in various device modes. Additionally, it empowers users by providing insights into the regular battery charging times, thus enabling them to determine the duration for which the device can be utilized to monitor patients. This knowledge proves invaluable for healthcare practitioners, as they can ensure uninterrupted monitoring while managing battery charging schedules effectively. Overall, this portable measurement system offers a promising solution for enhancing patient care and disease management in isolation treatment rooms.
Co-Authors Adi Purnama Bagus Alit Prasetyo Bambang Guruh Irianto Dodon Turianto Nugrahadi Herlistiono, Iwa Ovyawan Herteno, Rudy Luhur Bayuaji, Luhur Maghfiroh, Anita Miftahul Muliadi Purwitosari, Dyah Saragih, Triando Hamonangan Septiana, Silvi Dwi Sriyani Violina Syaifudin Syaifudin, Syaifudin Viddi Mardiansyah Zamzam, Yra Fatria