Current Research Focus
My interest in health is at the intersection of machine learning, behavioral sciences, and medicine. We commonly use modern machine learning methods with a focused effort to improve healthcare outcomes. I founded and lead the Rice Digital Health Initiative.
Rice is located across from the world’s biggest medical center, Texas Medical Center, and it gives us a major advantage in conducting high-impact research. In my Scalable Health Labs, we are developing methods to uncover behavior-biology causal pathways and develop novel speech and vision AI for health. With our clinical partners, we have access to unique retrospective data, and we regularly conduct novel prospective trials.
In addition to ongoing research directions in mental health and diabetes, we have new projects in cardiology, Parkinson’s, and DBS-based interventions for bipolar disorder. Several new positions are available in the broad area of AI for health.
Our lab commonly performs innovations in methods and their novel applications in clinical applications. Some recent examples (full publication list at my Google Scholar page)
Mental Health: We are developing foundations for a trans-diagnostic approach to psychiatry and establishing new bio-behavioral dimensions. Notably, we have developed quantitative methods to measure a person’s in-person socialization behaviors using a wrist-worn audio band and deep learning-based speech analysis pipelines to extract privacy-preserving socialization parameters like in-person network size, conversation behaviors, dyadic versus multi-party conversations, and social ambiance.
- “Towards objective, temporally resolved neurobehavioral predictors of emotional state,” Katherine Kabotyanski, et. al, Brain Stimulation, 2024.
- “Mobile sensing-based depression severity assessment in participants with heterogeneous mental health conditions,” Bishal Lamichhane, Nidal Moukaddam, Ashutosh Sabharwal, Scientific Reports, 2024.
- “RACER: An LLM-powered Methodology for Scalable Analysis of Semi-structured Mental Health Interviews,” SH Singh, K Jiang, K Bhasin, Ashutosh Sabharwal, N Moukaddam, AB Patel, Proceedings of the 1st Workshop on NLP for Science (NLP4Science), 2024.
- “ERSAM: Neural Architecture Search for Energy-Efficient and Real-Time Social Ambiance Measurement,” Chaojian Li, Wenwan Chen, Jiayi Yuan, Yingyan Celine Lin, Ashutosh Sabharwal, ICASSP 2023.
- “Modeling Suicidality with Multimodal Impulsivity Characterization in Participants with Mental Health Disorder,” Nidal Moukaddam, Bishal Lamichhane, Ramiro Salas, Wayne Goodman, Ashutosh Sabharwal, Behavioural neurology, 2023.
- “ECoNet: Estimating Everyday Conversational Network From Free-Living Audio for Mental Health Applications,” Bishal Lamichhane, Nidal Moukaddam, Ankit B Patel, Ashutosh Sabharwal, IEEE Pervasive Computing, 2022.
- “Privacy-preserving Social Ambiance Measure from Free-living Speech Associates with Chronic Depressive and Psychotic Disorders,” Wenwan Chen, Ashutosh Sabharwal, Erica Taylor, Ankit B Patel, Nidal Moukaddam, Frontiers in Psychiatry, 2021.
Diabetes: We have developed new bio-behavioral markers, new analysis methods and developed unique interventions as part of the NSF ERC PATHS-UP. Following is a snapshot of our 8+ years of research
- “Evaluating HbA1c-to-average glucose conversion with patient-specific kinetic models for diverse populations,” Sandra Emi Sato, Ashutosh Sabharwal, Wendy Bevier and David Kerr, Scientific Reports, 2024.
- “Multimodal digital phenotyping of diet, physical activity, and glycemia in Hispanic/Latino adults with or at risk of type 2 diabetes,” Amruta Pai, Rony Santiago, Namino Glantz, Wendy Bevier, Souptik Barua, Ashutosh Sabharwal, David Kerr, Nature Digital Medicine, 2024.
- “Patterns of timing and intensity of physical activity and HbA1c levels in hispanic/latino adults with or at risk of type 2 diabetes,” David Kerr, Mahsan Abbasi, Wendy Bevier, Namino Glantz, Arianna Larez, Ashutosh Sabharwal, Journal of Diabetes Science and Technology, 2024.
- “Clinical Characterization of Data‐Driven Diabetes Clusters of Pediatric Type 2 Diabetes,” Mahsan Abbasi, Mustafa Tosur, Marcela Astudillo, Ahmad Refaey, Ashutosh Sabharwal, Maria J Redondo, Pediatric Diabetes, 2023.
- “Discordance between Postprandial Plasma Glucose Measurement and Continuous Glucose Monitoring,” Souptik Barua, Raven A Wierzchowska-McNew, Nicolaas EP Deutz, Ashutosh Sabharwal, The American Journal of Clinical Nutrition, 2022.
- “The Northeast Glucose Drift: Stratification of Post-breakfast Dysglycemia among Predominantly Hispanic/Latino Adults at-risk or with Type 2 Diabetes,” Souptik Barua, Ashutosh Sabharwal, Namino Glantz, Casey Conneely, Arianna Larez, Wendy Bevier, David Kerr, LANCET EClinicalMedicine, 2022.
Past Research Directions
Pulmonary Conditions: Pulmonology was our first foray into health. That research led to our first spinoff, Cognita Labs, which has developed two FDA-approved products for the world’s first method to measure inhaler competence and a handheld lung function testing using the forced oscillometry technique. Our research on inhaler use taught us that human behavior should not be treated as a nuisance but as a necessary design requirement.
Retinal Imaging: We spent a few years developing a hand-held retinal imager, which could image the eye without dilation. That led to the second spinoff from the lab, OcuCheck.
Digital Gym: As an ambitious project, we wanted to track every exercise performed in the Rice Gym, automatically. The project generated many ideas and one of them led to our third spinoff, CherryPick.