580 California St., Suite 400
San Francisco, CA, 94104
Key research themes
1. How do large-scale multi-modal imaging-genomic datasets enable detailed genotype-phenotype associations and precision medicine?
This research theme centers on leveraging extensive imaging and genomic datasets to elucidate intricate genotype-phenotype relationships. The convergence of large-scale imaging cohorts with genomic profiling facilitates a systems-level understanding of disease mechanisms, enabling precision medicine approaches that tailor diagnosis and treatment to individual molecular and phenotypic profiles. It matters because such integrative datasets provide statistical power to detect subtle associations, enable identification of novel imaging biomarkers linked to genomics, and support AI-driven analyses that exceed the scope of traditional small-scale studies.
Key finding: The UK Biobank's imaging enhancement integrates extensive multi-modal imaging (brain, cardiac, abdominal MRI, DXA, ultrasound) with comprehensive genetic and phenotypic data on 100,000 participants, enabling population-scale... Read more
Key finding: By combining high-resolution digital whole slide images with genomic and clinical data from The Cancer Genome Atlas, the study demonstrates how advanced imaging feature extraction coupled with genomics can identify... Read more
Key finding: Imaging-AMARETTO software integrates multiomics, radiography, and histopathology data from large cancer cohorts to model regulatory networks and link them with imaging biomarkers predicting clinical outcomes. Application to... Read more
Key finding: PRISM extends data resources like the Cancer Imaging Archive by managing and curating integrated radiology, pathology images, clinical data, and radiomic/pathomic features. The modular and scalable platform addresses the... Read more
Key finding: This paper highlights how molecular imaging modalities such as PET, SPECT, MRI, and ultrasound provide in vivo characterization of molecular events crucial for personalized medicine, especially in oncology. Molecular imaging... Read more
2. What technological advancements in molecular and multi-scale imaging modalities enhance the visualization and quantification of genetic and molecular information in situ?
This theme investigates the development of cutting-edge imaging modalities that push the resolution and molecular specificity boundaries, allowing single-cell and subcellular analysis of genetic and molecular features. Techniques including super-resolution microscopy, optical DNA mapping, SIMS imaging, and integrated imaging mass spectrometry with microscopy are explored. These methods provide unprecedented views into chromatin organization, spatial genomics, molecular tissue mapping, and cell-biology relevant molecular distributions, which are essential for detailed mechanistic insights and for correlating molecular events with imaging phenotypes.
Key finding: Introduces Oligopaint probes combined with single-molecule super-resolution microscopy enabling visualization of single-copy genomic regions with haplotype-specific resolution via SNP-targeted hybridization. Demonstrates... Read more
Key finding: Reviews optical mapping techniques for linearized DNA molecules spanning tens to megabase pairs, which allow direct visualization of sequence motifs, epigenetic marks, and molecular barcodes via fluorescence microscopy.... Read more
Key finding: Presents an image fusion approach integrating mass spectrometry imaging (IMS) with high-resolution optical microscopy via multivariate regression to predict molecular distributions at spatial resolutions exceeding IMS... Read more
Key finding: Reviews secondary ion mass spectrometry (SIMS) applications with subcellular spatial resolution enabling multiplexed, label-free chemical imaging of lipids, metabolites, and proteins in neurobiology and cell biology.... Read more
Key finding: Develops optimized DNA-PAINT super-resolution imaging protocols that increase acquisition speed by 10-fold through rational sequence and buffer optimization, while maintaining spatial resolution below 5 nm. This advancement... Read more
3. How can computational and informatics platforms facilitate integration, analysis, and interpretation of complex imaging-genomic data in precision medicine?
This theme explores the design and implementation of computational infrastructures, software tools, and analytic frameworks that enable efficient management, integration, and extraction of knowledge from heterogeneous imaging, multiomics, and clinical data. It emphasizes challenges in data harmonization, open-source accessibility, scalability, multi-modal fusion, and machine learning applications for phenotype-genotype correlation and predictive modeling. Such platforms are critical to advancing personalized diagnostics, radiogenomics, and image-based biomarker discovery.
Key finding: Presents MIRA, an open-source, web-accessible molecular imaging repository and analysis platform that supports integration of diverse imaging modalities, bioinformatics, and clinical data across multicenter collaborations.... Read more
Key finding: Reviews a spectrum of computational approaches including image reconstruction, enhancement, multi-modal image processing, and machine learning for molecular imaging modalities like PET and SPECT. Highlights their roles in... Read more
Key finding: Introduces MITK Phenotyping, an open-source radiomics framework that implements standardized imaging feature extraction per IBSI guidelines with multiple user interfaces and programming language support. Facilitates... Read more
Key finding: Demonstrates a deep learning based approach integrating patient facial photographs to assist clinical exome interpretation, achieving 99% top-10 accuracy in disease-gene prioritization for 679 individuals with monogenic... Read more
Key finding: Reviews the integration of radiomics and genomics termed radiogenomics, outlining classical and deep learning radiomic pipelines linking imaging features to tumor genomics for improved diagnostics, prognostics and treatment... Read more