Dr Katharina Brosch
I build computational models of mental health, from how early environments shape brain development to how speech may predict outcomes in psychiatric illness.
Skills
Python · scikit-learn · R · machine learning · NLP / digital phenotyping · large-scale longitudinal data · neuroimaging (SPM, FSL, CAT12) · statistical modeling
About
I’m a neuroscientist and psychologist working on a single problem: most psychiatric care begins only after illness has emerged. My research aims to move that line earlier, toward earlier identification and prevention.
As a postdoctoral researcher at the Feinstein Institutes for Medical Research (Northwell Health), I build computational models of mental health across the lifespan. On one side, I use large-scale data, including the ABCD Study of over 10,000 adolescents, to model how socioeconomic conditions and early adversity relate to brain development. On the other, I work on speech-based markers in schizophrenia, quantifying acoustic and linguistic features of speech with the aim of predicting social functioning and symptom course.
What ties these together is a focus on scalable, objective markers and on modifiable targets, the factors we can actually act on. I pay particular attention to protective factors, not just risk, because understanding what keeps people well matters as much as understanding what makes them ill. I work primarily in Python and R, and have led research collaborations across Germany and the U.S.
How I work
I care about methods that hold up and results that translate. That means careful validation, honesty about what the data can and can’t show, and a focus on findings that can reach real clinical or product decisions, not just publications.
Dr Katharina Brosch 