My PhD focuses on the application of multimodal machine learning for prognostics and diagnostics in mood disorders. For example, the prediction of rehospitalization in inpatient unipolar depressive samples using a combination of clinical, structural imaging, genetic, cardiovascular and blood-biomarker predictors, as well as the clinical and genetic prediction of Lithium response in bipolar samples. More broadly, I have an interest in translation, commercialization, and industry engagement, as demonstrated by my time spent at Spring Health in New York city. Spring Health is a US healthcare company born out of Yale’s psychiatry department that uses machine learning and analytics to personalize and decrease the time needed to achieve adequate mental health care. In my time at this company, I analyzed electronic medical records related to the off-label use of intravenous ketamine at different psychiatric clinics across the united states.
Regarding my current viewpoints, I believe that academia is burdened by bureaucratic inefficiencies, a misguided focus on inference over translation/prediction, and a poor appetite for risk. Resultantly, leading to suboptimal and slow to market treatments for patients as well “unintentional” patient harms that arise from “not-acting” and avoiding risk. Whilst some of these conditions are unavoidable and arise as a negative, yet, necessary externality of ensuring patient safety, I believe that a more optimal point of balance is attainable and can be facilitated through a transparent industry/academia relationship that combines the efficiency of industry with the intellectual capital of academia. Within the IMNIS program, I hope to better refine these developing thoughts and begin to develop a symbiotic relationship with those in industry.