Measuring speech biomarkers of depression in daily life: patient-centered validation of a novel tool in a community sample

About This Grant

PROJECT SUMMARY To address challenges in scaling speech-based biomarker (SBB) research and diagnostics, the applicant developed and piloted a mobile app for daily SBB capture (“Fabla”), funded by the Georgia CTSA (UL1TR002378). In alignment with NIMH priorities of identifying biomarkers of mental illness and innovations in digital health technologies, this K01 project will validate and optimize Fabla for assessing SBBs associated with major depressive disorder (MDD) in a community sample. Candidate: Dr. Kaplan is a clinical psychologist with an extensive background in ecological assessment, acquired through doctoral training and an F32 NRSA fellowship. An Assistant Professor at Emory University School of Medicine, her career aim is to lead an R01- funded program of research that develops and implements mobile SBB technologies for screening, treatment, and relapse monitoring of depression and other mental health concerns. Training: Dr. Kaplan’s career goals require additional training in technology-specific domains that are crucial for translational SBB research: technology co-design, mobile app product management and specification (including clinical ethics concerns), and expanded training in biostatistics, including paralinguistic feature analysis. Accordingly, this K01 leverages multidisciplinary training in Emory’s departments of Psychiatry, Biomedical Engineering, and Bioinformatics. Training includes formal coursework, individualized training with mentors, and structured career development support. Mentoring: This K01 adopts a co-primary mentorship model and is jointly sponsored by Wilbur Lam, MD, PhD (Professor of Biomedical Engineering, Vice Provost of Entrepreneurship, and a leader in mobile diagnostics) and Barbara Rothbaum, PhD (Professor of Psychiatry with expertise in mood disorders and mental health technology development). Co-mentors and significant contributors bring expertise in co-design (Gari Clifford, DPhil), paralinguistic analytics (Gari Clifford, DPhil and Shrikanth Narayanan, PhD), machine learning (Anant Madabhushi, MD), MDD (Boadie Dunlop, MD, MSc), and recruiting and engaging older adults (Madeleine Hackney, PhD) Research: Leveraging established research partnerships with multiple Emory clinics and active longitudinal community studies, this project recruits n=300 individuals (n=150 with MDD, n=150 psychiatrically healthy) to complete a brief voice diary using the Fabla app for 10 days. The project aims to (1) optimize Fabla for MDD screening using co-design methodology, (2) characterize Fabla-assessed SBBs and their preliminary convergent validity with published norms, and (3) evaluate predictive validity of Fabla- assessed SBBs for predicting MDD using machine learning. Products include six planned manuscripts; a structured data repository to facilitate norm development; updates to the Fabla app; and R34 and R01 grants submitted by the applicant. Environment: The candidate receives strong institutional support from the Department of Family and Preventive Medicine. The training environment includes multiple Emory/Georgia Tech health technology institutes and Emory Outpatient Psychiatry, the site for the applicant’s planned R01.