Research

Computational Medical Imaging Laboratory (CMIL) develops novel computational methods to study and understand tissue micro-anatomy using digital brightfield microscopy histology and cutting-edge spatial-omics data. We also conduct multi-modal AI (MMAI) model development by integrating imaging and omics data with clinical EHR and unstructured notes data, develop standardized OMOP multimodal AI data infrastructure for AI readiness, implement federated architecture for privacy and security and trustworthy AI, and develop end-user tool via codesign based on usability and ethics study, and disseminate our tools following gold-standard science concept. Methods developed by CMIL team inform clinical diagnostics and allow for the study of the fundamentals of biological systems. Currently, major focus of CMIL is diabetic kidney disease, kidney transplant, and normal reference kidneys in humans.

Schematic of our MMAI concept, for integrating multi-modal data involving pathology, radiology, EHR, and clinical text data, OMOP data standardization, multi-moda model development, privacy and security preserved federated deployment of the model, end-user tool development via system co-design using usability and ethics studies.
Schematic of our MMAI concept, for integrating multi-modal data involving pathology, radiology, EHR, and clinical text data, OMOP data standardization, multi-moda model development, privacy and security preserved federated deployment of the model, end-user tool development via system co-design using usability and ethics studies.

Interests

  • Digital & Computational Pathology
  • Microscopy Image Analysis
  • Image Processing
  • Multi-Omics Data Fusion
  • Multi-Modal AI
  • End-User Cloud Tool Development
  • Usability, Ethical AI, & System Co-Design