Faculty in the Statistics in Imaging Hub develop statistical methods for analyzing complex biomedical imaging data, including radiologic, pathologic, and advanced MRI modalities. Their work focuses on quantitative feature extraction, spatial and functional modeling, and longitudinal analysis of high-dimensional images.
Investigators collaborate with imaging scientists and domain experts to design rigorous, reproducible frameworks for radiomics and quantitative pathology, enabling robust characterization of imaging-derived phenotypes across scales. By advancing principled statistical approaches for image-based data, this Hub enhances interpretability and reliability in imaging studies of cancer and related diseases.
Hub Faculty
Jing Ma
Microbiome and Multi-Omics Integration · Network and Graphical Models · Brain Connectivity/Imaging Data
Jing Ma develops statistical machine learning and network analysis methods for complex, high-dimensional biomedical data, with applications spanning microbiome studies, systems biology, and imaging-derived data. Her work on graphical models and Granger causality networks advances inference for brain connectivity and other spatially structured imaging data. By integrating network biology with image-based data, she develops methods to identify latent structure, characterize biological pathways, and relate imaging phenotypes to underlying molecular processes through her work in the Ma Lab. Her research emphasizes robust, reproducible software tools that support scalable analysis across data modalities.
Wei Sun
Pathology Imaging and Deep Learning · Spatial Transcriptomics and Omics Integration · Tumor Microenvironment Modeling
Wei Sun's lab evelops statistical and machine learning frameworks that integrate imaging and omics data, with a focus on deep learning methods for pathology image analysis and joint modeling of H&E images with spatial proteomic data. His work advances quantitative characterization of the tumor microenvironment by modeling cellular architecture, spatial organization, and molecular heterogeneity. Through contributions to graph neural networks and multi-modal data integration, his research establishes imaging as a critical quantitative layer for studying cancer biology and immune–tumor interactions.