Machine Learning Prediction of SSRI Response Using fMRI Connectivity
CWSF · 2026 Disease & Illness Bronze Medal
Overview
When someone is depressed, doctors prescribe antidepressant medication, but it only works for about half of patients, meaning the other half can wait months on ineffective treatment. We developed a machine learning model that predicts whether a medication will be effective by analyzing brain connectivity patterns before treatment begins. Using a synthesized dataset of 138 virtual patients designed to reflect real clinical data, our model achieved 88% accuracy in predicting treatment success, compared to 54% accuracy using traditional clinical information such as age and symptom severity. We also built a web tool where this prediction model can be tested in real time. If validated on real patient data, this approach could help millions of people with depression get the right treatment faster, reducing unnecessary suffering and healthcare costs.
Awards (2)
- Bronze Medal
- Selected for CWSF 2026
Competition history
- CWSF 2026
Related projects
ISEF · 2025
A Novel Machine Learning Method and Drug for the Diagnosis and Treatment of Depression
ISEF · 2022
Predicting Onset of Depressive Disorder Using Machine Learning
ISEF · 2019
A Novel Noninvasive and Inexpensive Biomarker for Diagnosing Major Depressive Disorder (MDD): Using Machine Learning Model in silico and Drosophila melanogaster Model in vivo
ISEF · 2026
Early Diagnosis of Schizophrenia Through an Interpretable CNN–Transformer Model of fMRI Connectivity
Closest projects by meaning, across every fair and year in the corpus.