Research Statement

Introduction

I am Associate Professor (with tenure) at The University of Texas at MD Anderson Cancer Center, where I joined in June 2021. Previously, I was Assistant Professor at Emory University. I have published over 62 peer-reviewed journal articles, and several other manuscripts are under revision or submitted. These articles have been published in top-tier journals such as The Journal of the American Statistical Association, Biometrics, NeuroImage, Human Brain Mapping, Biostatistics, Bayesian Analysis, PLoS One, Scientific Reports, The Journal of Machine Learning Research, and so on.

I am considered a thought leader in developing statistical methods for high-dimensional medical imaging data, with a goal to tackle mental health disorders such as PTSD or model the progression of neurodegenerative diseases. I specialize in the integrative analysis of large scale multi-site imaging data that can often span multiple modalities (structural and functional) or experiments, while also developing tools for combining imaging and omics information for a more holistic characterization of the disease landscape.

I have published several novel methodological papers to address the critical question of measurement errors in high-dimensional biomedical datasets that has lead to considerable improvements in predictive accuracy and higher power for biomarker discovery in neuroimaging and -omics studies. My work on Bayesian tensor modeling for spatial functional data has had a strong impact in neuroimaging literature, and my recent work on spatially-aware harmonization methods for large scale multi-site studies has garnered strong interest as evident from invited talks in several national universities.

Impact

The impact of my work is evident from media attention in platforms such as Science Trends and several best paper awards received in national conferences for articles where I am the first/corresponding/senior author. In addition, The Institute of Data Science in Oncology at MD Anderson has expressed strong interest in adopting the spatially-aware neuroimaging harmonization pipeline developed by my lab.

I have a successful track record of obtaining extramural funding as demonstrated by several awards, including an R01 research grant from NIMH (Role: PI, 2019–2025), another R01 research grant from NIA (Role: PI, 2021–2026), and several other grants where I have served as a co-investigator. Moreover, I am the recipient of numerous awards, including the prestigious 2019 Young Statistical Scientist Award awarded by the International Indian Statistical Association, and Elected Member of the International Statistical Institute (ISI) that demonstrates the visibility and impact of my research and teaching.

I have rich interdisciplinary and collaborative research experience with biomedical and computer science investigators and have served as a co-investigator on several collaborative grants. My leadership is visible through various roles, as demonstrated by serving as the Director for the Data Analytics & Biostatistics Core in the School of Medicine at Emory University, and serving as an Editorial Board member of prestigious statistics journals such as Biometrics and the official journal on statistical methods in imaging published by The American Statistical Association titled Statistics and Data Science in Imaging.

My CV will list a strong track record of service activities both in the professional societies as well as within my current institute (MD Anderson). Finally, I deeply value opportunities to mentor and train the next generation of scientists, as reflected in the successful accomplishments of my graduate students and postdoctoral fellows. While teaching is not a formal requirement at MD Anderson, my experience teaching both before and after joining the institution demonstrates a strong commitment to education and fostering academic growth.

Research Summary and Future Directions

My research spans several methodological and collaborative areas, emphasizing innovation and practical impact. Methodologically, I focus on (1) Bayesian network analysis; (2) Bayesian functional data analysis; (3) Bayesian tensor modeling; (4) Bayesian functional data clustering; (5) biomedical data analysis with measurement error; (6) predictive modeling incorporating prior knowledge; and (7) integrative analysis of multi-modal and multi-view data. These approaches have significantly influenced (1) modeling the progression of mental and neurodegenerative disorders; (2) early detection and risk prediction; (3) biomarker discovery via inference; and (4) advancing personalized medicine. Looking ahead, I aim to further develop (1) interpretable AI techniques and digital twins for more precise modeling; and (2) tailor spatially-aware methods for multi-center trials to enhance reproducibility and clinical relevance. These efforts will ensure that my research continues to bridge methodological advancements with transformative biomedical applications.