Title : Raman scattering physics and AI-based spectral analysis for earlier diagnosis of Parkinson’s disease
Abstract:
Parkinson’s Disease (PD) is a progressive neurodegenerative disorder characterized by dopaminergic neuronal loss and α-synuclein aggregation, where early diagnosis remains a fundamental challenge due to the absence of sensitive, non-invasive biomarkers. From a physical perspective, Raman spectroscopy provides a powerful probe of molecular structure through inelastic light–matter interactions, enabling the detection of vibrational fingerprints associated with biochemical alterations in complex biological systems. In this work, we present a physics-driven diagnostic framework that integrates Raman spectroscopy with advanced artificial intelligence (AI) algorithms for the identification and monitoring of PD. The methodology is grounded in the quantitative analysis of Raman scattering signals, where spectral intensity, molecular concentration, and scattering cross-sections collectively encode pathological information. Acquired spectra undergo systematic preprocessing, followed by dimensionality reduction and classification using machine learning and deep learning architectures. The proposed system demonstrates high diagnostic performance, with reported accuracies exceeding 90% and strong sensitivity and specificity across multiple biological matrices, including biofluids and neural tissues. Importantly, the approach enables the detection of subtle physicochemical changes related to protein conformational transitions, lipid remodeling, and oxidative stress. This study highlights the potential of combining the fundamental physics of vibrational spectroscopy with AI-driven data analysis to establish a non-invasive, high-resolution diagnostic platform. The proposed framework paves the way for early detection, real-time monitoring, and precision medicine in Parkinson’s disease.
