This study aims to investigate how multimodal data can be used to predict and understand different antidepressant medication/treatment response in individuals with Major Depressive Disorder (MDD), leveraging UK Biobank’s extensive resources.
Research questions:
1. To what extent can brain imaging data (functional MRI), demographic information and electronic health records be used with machine learning approaches to predict antidepressant treatment response in individuals with MDD?
2. What are the most effective strategies for representing and integrating heterogeneous data sources to develop robust machine learning pipelines for treatment response prediction?
3. How does stratification of the patient cohort based on clinical, demographic, or diagnoses information influence predictive model performance and the identification of clinically meaningful markers of treatment response?
Objectives:
a) To quantify antidepressant treatment response in individuals with MDD
b) To investigate whether fMRI and other scales are associated with baseline symptom severity and treatment outcomes.
Scientific rationale:
MDD is a leading global cause of disability, yet treatment efficacy is highly variable in real-world settings. Clinical trials often lack generalisability to broader populations, underscoring the need for large-scale, naturalistic evidence on treatment outcomes and predictors of response. UK Biobank provides a unique opportunity to address this gap by combining self-reported mental health data, prescription records, hospital and primary care data, demographic information, and brain imaging-derived phenotypes. This research will help identify factors associated with treatment response and inform the development of personalised treatment strategies, with the ultimate goal of improving outcomes for individuals with MDD.