research

My research spans clinical NLP, speech and language technologies for healthcare, and computational epidemiology. I am interested in building interpretable, equitable AI systems for health applications.

Publications & Submissions

Interpretable Phonological–Semantic Dissociation Scoring for Automated svPPA Subtyping

Sarika Pasumarthy, Minggang Li, Jiachen Lian, Chenxu Guo, Emma Xuedi Yang, Lynn Kurteff, Zoe Ezzes, Willa Keegan-Rodewald, Jet M.J. Vonk, Siddarth Ramkrishnan, Giada Antonicelli, Zachary A. Miller, Maria Luisa Gorno-Tempini, Gopala Anumanchipalli

We develop an interpretable scoring framework for automated subtyping of semantic variant primary progressive aphasia (svPPA), dissociating phonological and semantic deficits from naturalistic speech.

S-MARC: Causal Streaming Reasoning for Full-Duplex Conversational Behavior Modeling

Dingkun Zhou, Shuchang Pan, Siddharth Banerjee, Tiantian Feng, Sarika Pasumarthy, Dhruv Hebbar, Siddhant Patel, Zeyi Austin Li, Kan Jen Cheng, Sanay Bordia, Krish Patel, Akshaj Gupta, Tingle Li, Gopala Anumanchipalli, Jiachen Lian

S-MARC introduces a causal streaming reasoning framework for modeling full-duplex conversational behavior, enabling real-time turn-taking and backchanneling decisions.

Heterogeneity across Race and Ethnicity for Menopause Onset

Sarika Pasumarthy, Maggie Hurley, Irene Y. Chen, Nitya Thakkar, Monica Agrawal, Yulin Hswen

Submitted to npj Women’s Health (Nature Portfolio)

Using the NIH All of Us Research Program Controlled Tier dataset (v8, n = 11,306), we analyze age at first menopause diagnosis across three progressively refined cohorts, finding that Asian & Pacific Islander and Indigenous/Other individuals experience significantly earlier menopause onset than White individuals after covariate adjustment.

Research Groups

Berkeley Speech Group — Prof. Gopala Anumanchipalli, UC Berkeley EECS Speech and language technologies for neurocognitive disorder detection.

Computational Precision Health — Prof. Irene Chen & Prof. Yulin Hswen, UC Berkeley & UCSF Large-scale EHR analysis, sex differences in multimorbidity, and reproductive health equity.