Identifying diagnostic biomarkers in functional motor disorders through multimodal behavioral, neurophysiological,...

Area of Science:

• Neurology
• Neuroscience
• Biomarker Discovery

Background:

• Functional motor disorders (FMDs) are common, disabling neurological conditions.
• Diagnosis can be challenging due to heterogeneity and lack of reliable biomarkers.

Purpose of the Study:

• To identify multimodal biomarkers for distinguishing FMDs from healthy controls (HCs).
• To leverage machine learning for objective diagnostic tools in FMDs.

Main Methods:

• A multicenter cross-sectional study involving 75 adults with FMDs and 75 age/sex-matched HCs.
• Standardized behavioral, neurophysiological, and brain MRI assessments.
• A Random Forest classifier trained on multimodal features for diagnostic prediction.

Main Results:

• Key biomarkers include dual-task effect scores, gait speed, vDMN and basal ganglia connectivity, blink reflex area, and DNIC-to-baseline amplitude ratios.
• The Random Forest classifier achieved high performance: 85.0% accuracy, 86.1% specificity, and 0.921 AUC-ROC.

Conclusions:

• Multimodal markers (motor, neuroimaging, neurophysiological) effectively differentiate FMDs from HCs.
• These findings address the need for objective tools, supporting more confident FMD diagnosis.