The primary focus of my research has been developing new artificial intelligence (AI) approaches including image analytics, radiomics, and machine learning schemes; applied to problems in computer-aided diagnosis & detection, disease characterization, as well as quantitative evaluation of response to treatment; in gastrointestinal cancers and digestive diseases. The key innovation here lies in designing unique AI tools that can capture biologically relevant and clinically intuitive measurements from routinely acquired imaging (MRI, CT, PET) or digitized images of tissue specimens. Critically, to “unlock” embedded information captured by these modalities, our AI tools inform and enrich imaging measurements with spatially resolved molecular, serum, or pathologic information. This in turn enables cross-scale association between imaging, pathology, and -omics data modalities towards building more accurate and generalizable computational imaging predictors that offer improved risk stratification, disease modeling, and biological quantitation in vivo.
I have authored nearly 50 peer-reviewed journal publications, over 100 conference papers & abstracts, 1 book chapter, as well as delivered over 60 invited talks and panel discussions both in the US and abroad. I have 10 issued patents in the areas of medical image analysis, computer-aided diagnosis, and pattern recognition. I am an Associate Editor for 3 leading medical imaging journals, serve on Program Committees for 3 major medical image analytics conferences, and have been elected to Senior Member in the National Academy of Inventors, the IEEE, and the SPIE. In 2023, I was selected for a Fulbright Specialist Award, as a Notable in Education Leadership by Crain’s Cleveland, and for the SIIM Imaging Informatics Innovator Award. My research has been continuously funded since 2016 through the DOD/CDMRP, the NIH (NCI, NIDDK, NINR, NHLBI), as well as the State of Ohio.
Research Information
Research Interests
The primary focus of my lab is developing novel medical image analysis and machine learning tools for imaging data, through spatial correlation and cross-linking against pathology or molecular data. Applications of our tools are being examined in: (a) decision support for treatment (e.g. choice of therapy), (b) targeting therapeutic procedures (e.g. guiding ablation, radiotherapy, surgery), and (c) biological quantitation for treatment response characterization in vivo. This multi-disciplinary, multi-pronged approach is being applied to colorectal, renal, and prostate cancers, as well as digestive diseases.