Multimodal approach for early diagnosis of Parkinson's disease using PET imaging, tremor detection, and machine learning.

Parkinson's disease (PD) is a progressive neurodegenerative disorder that presents diagnostic challenges, particularly in its early stages. This study proposes a multimodal approach to PD classification that integrates neurological imaging, motor symptom analysis, and non-motor clinical features. Dopamine depletion, a core biomarker of PD, is assessed using PET imaging, where active brain regions are quantified through color segmentation and image processing. A reduction in the active area correlates with disease progression. Tremor detection is performed using the Hough Transform algorithm applied to line-drawing tests, effectively identifying motor irregularities. Non-motor features are analyzed using a publicly available dataset, and the XGBoost algorithm achieves a classification accuracy exceeding 95.42%. The combined approach demonstrates high potential for early, accurate, and interpretable PD diagnosis.