A Knowledge-Driven Framework Discovers Brain ACtivation-Transition-Spectrum (ACTS) Features for Parkinson's Disease.

Area of Science:

• Neuroscience
• Biomedical Engineering
• Medical Imaging

Background:

• Conventional Parkinson's disease (PD) treatment assessment relies on subjective human evaluation, leading to variability.
• Dopaminergic therapy is effective for PD, but objective, quantifiable assessment methods are needed.

Purpose of the Study:

• To develop a knowledge-driven framework using brain ACtivation-Transition-Spectrum (ACTS) features for objective assessment of dopaminergic therapy in PD.
• To validate the efficacy of ACTS features in quantifying treatment-induced changes in brain activity.

Main Methods:

• Functional near-infrared spectroscopy (fNIRS) recorded brain activity in 51 PD patients during OFF and ON dopaminergic medication states.
• ACTS features were derived from regional activation, hemodynamic state transitions, and functional connectivity graph spectrum.
• A feature selection algorithm combined recursive feature elimination and graph spectrum analysis, followed by linear discriminant analysis.

Main Results:

• The proposed ACTS feature framework reduced misclassification probability in treatment evaluation from 42% to 16%.
• The method demonstrated superior performance compared to existing fNIRS-based approaches for PD treatment assessment.
• Identified ACTS features show potential as functional biomarkers for objective therapy evaluation.

Conclusions:

• The developed framework offers a quantified and objective approach for assessing dopaminergic treatment in Parkinson's disease.
• Brain ACTS features represent promising novel biomarkers for reliable and objective evaluation of PD treatment efficacy.