INDIVIDUALIZED BRAIN DISEASE PREDICTION METHOD AND SYSTEM BASED ON MAGNETIC RESONANCE IMAGING

摘要

An individualized brain disease prediction method and system (50) based on magnetic resonance imaging. The method comprises: Step 1, acquiring a brain magnetic resonance image of a psychiatric patient (S1); Step 2, performing denoising and dimensionality reduction on the brain magnetic resonance image of the patient (S2); Step 3, using an ReliefF algorithm to perform feature selection (S3); Step 4, using a spatial cluster analysis method to adaptively obtain a spatial brain region (S4); Step 5, using a correlation-based feature selection algorithm to remove a redundant feature and obtain a best feature subset (S5); Step 6, performing, on the basis of the best feature subset, a multiple linear regression analysis to identify a potential biomarker (S6). The system (50) comprises: an acquiring module (51), a pre-processing module (52), a feature selection module (53), a clustering module (54), a screening module (55), and an identification module (56). The method and system (50) can quantify, on the basis of various image data of different modal types, a feature of interest in a psychiatric disease and perform individualized prediction, thereby enhancing identification of potential pathogenesis of a psychiatric disease.