Arvind Rao, PhD

Research theme: my group’s research has developed methodologies for the analysis and integrative interpretation of spatially-based, multi-modal, omics datasets (e.g: imaging, genomics, spatial profiling) for improved disease understanding and clinical decision making. Our work spans machine learning and statistical modeling for radiomics, digital pathology, predictive modeling, genomic data integration, and most recently spatial immunoprofiling and spatial transcriptomics.

Research:

a. Biomedical Image Informatics & Spatial Biology: In order to quantify the phenotypic aspects of disease, their relationships with outcome, and their genetic context, we have developed methods for the analysis of histopathology and radiology images, focusing on tumor heterogeneity. One direction of my group is to develop image analysis tools to delineate tumor image features from radiology data and, along with underlying genomic measurements, to develop AI-based predictive models to relate them to outcomes in low grade gliomas. Further, we have also investigated methodologies to link tumor imaging, genetics, and immune status in disease. Along these lines, we have embarked on measuring immune contexture in the tumor microenvironment, developing methods for the inference of the spatial architecture and patterns of infiltration of immune cells, based on multiplex-IHC and spatial transcriptomics technologies. Further, we have also developed methods for the analysis of multiparametric magnetic resonance imaging (MR) datasets in radiation oncology

b. Artificial Intelligence (AI) methods for Medicine: Our expertise is evidenced by two streams of work: (1) Developing statistical and machine learning (ML) methods to analyze and interpret multi-modal image/genomic data; (2) application of AI/ML formalisms for predictive models based on genomics, radiology and pathology imaging. These are used to assess disease stage, suitability for specific drug targets, molecular therapies & immunotherapies. More recently, we have been working on problems of bias, uncertainty quantification as well as auditing of AI solutions (in radiology and pathology image informatics) in healthcare.