Domain-Specific Prediction of Clinical Progression in Parkinson's Disease Using the Mosaic Approach

Area of Science:

• Neurology
• Biomarker Discovery
• Neuroimaging

Background:

• Parkinson's disease (PD) presents with highly individualized symptoms.
• Cortical disease burden, assessed via MRI, may correlate with clinical heterogeneity.
• Personalized PD care may benefit from domain-specific neurological disability assessments.

Purpose of the Study:

• To evaluate the Mosaic Approach (MAP) for quantifying individual cortical disease burden.
• To assess MAP's potential in predicting domain-specific clinical progression in PD.
• To determine if an extremity-specific MAP improves prediction compared to a whole-brain MAP.

Main Methods:

• Utilized MRI and clinical data from 135 recently diagnosed PD patients (Parkinson's Progression Markers Initiative).
• Defined an extremity-specific motor score and identified corresponding cortical regions for a restricted MAP.
• Contrasted explanatory power of restricted vs. unrestricted MAP for motor and cognitive functions, using support vector machines for progression prediction.

Main Results:

• The extremity-specific MAP showed higher explanatory power for extremity-specific motor function than the unrestricted MAP.
• The unrestricted MAP better explained general motor function.
• No associations were found between MAP and cognitive function.
• Extremity-specific MAP predicted extremity-specific motor progression over 1 and 3 years.

Conclusions:

• The MAP framework enables domain-specific prediction of PD progression.
• This approach can inform machine learning models for personalized PD patient care.
• Customized, high-resolution assessment of cortical pathology shows promise as a PD biomarker.