NeuroSense: Voice-Based Screening for Parkinson’s Disease Using Classical ML on Dysphonia Biomarkers

Author(s)	Mr. Vishay Agarwal
Country	United States
Abstract	While prior studies have explored voice-based detection of Parkinson’s disease (PD), most rely on large neural networks, continuous speech, or black-box models that are difficult to interpret, reproduce, or deploy in low-resource settings. NeuroSense takes a fundamentally different approach: it demonstrates that classical machine learning on a compact set of dysphonia biomarkers—jitter, shimmer, harmonic-to-noise ratios, and nonlinear dynamical features—can achieve state-of-the-art discrimination (~92% test accuracy) using only sustained vowel phonations. By prioritizing interpretability, reproducibility, and lightweight deployment, NeuroSense challenges the prevailing assumption that effective PD voice screening requires deep learning or complex audio pipelines. This pipeline not only produces physiologically meaningful feature importance but also provides a fully auditable, clinically informed baseline that can be applied across research and early screening contexts. NeuroSense sets a new standard for transparent, accessible, and high-performing voice-based PD detection.
Keywords	Parkinson’s disease, dysphonia, acoustic biomarkers, interpretable ML, reproducible AI, vocal fold dynamics.
Field	Biology > Medical / Physiology
Published In	Volume 7, Issue 2, March-April 2026
Published On	2026-03-13
DOI	https://doi.org/10.63363/aijfr.2026.v07i02.3711

About AIJFR Fees & Payment Current Issue Publication Archive	Submit Research Paper Track Submission Status Publication Guidelines Publication Ethics Peer Review & Plagiarism	Join as a Reviewer Editors & Reviewers Reviewer Referral Program Get Reviewer Membership Certi.	Website/Journal Policies Usage Policy Content Policies Privacy Policy

Contact Us		+91-9898-79-39-59	editor@aijfr.com

Advanced International Journal for Research