Interpretable ML predicts Parkinson's severity using MRI and fMRI

Researchers have developed an interpretable machine learning model capable of predicting Parkinson's disease motor severity using a combination of QSM MRI and multiband multiecho fMRI features. The study found that imaging-only models showed significant predictive power, with combined QSM and clinical variables explaining 45.4% of the variance in motor severity. Specifically, selected QSM and clinical features were most effective, predicting motor severity within a 5-point range for 75.0% of participants, with key features highlighted in the cerebellum, thalamus, striatum, insula, and motor cortex. AI

IMPACT This research demonstrates the potential for interpretable AI models to improve the objective assessment of neurological conditions like Parkinson's disease.

RANK_REASON The cluster contains a research paper detailing a new machine learning model for medical diagnosis. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• elastic net regularization
• MDS-UPDRS Part III
• multiband multiecho fMRI
• Parkinson's disease
• QSM MRI
• random forest
• Shap
• Support Vector Regression
• XGBoost

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →