Outline of Proposed Research
This project will investigate the associations of the environment with the brain structure in health and disease, and account for this in imaging biomarkers such as treatment-response in psychosis.
Research Questions:
How do shared environmental characteristics influence brain structure in health and psychosis?
Does environmental adversity reduce the discriminative capacity of imaging biomarkers for clinically relevant subgroups (e.g., treatment responders vs. non-responders)?
Can incorporating environmental data into predictive algorithms improve their performance and generalizability across contexts?
Objectives:
Investigate effects of migration on brain structure in healthy individuals and those reporting psychotic-like experiences (UK Biobank).
Develop and test machine learning models predicting treatment response in first-episode psychosis from structural MRI, assessing performance across environmental contexts.
Compare strategies for integrating environmental factors into prediction pipelines to enhance reliability and reduce health inequities.
Scientific Rationale:
Most neuroimaging biomarker research assumes brain structure is context-independent, limiting replication and applicability in diverse settings. Evidence shows macro-environmental adversity shapes brain anatomy in healthy individuals, potentially shifting the baseline for pathological conditions. Ignoring these effects risks developing biomarkers that underperform in vulnerable populations, exacerbating health disparities. By integrating environmental measures into biomarker development, this project aims to produce more robust, equitable tools for precision psychiatry, with particular relevance for high-inequality regions like Latin America.