Parkinson’s Disease Detection From Online Handwriting Based on Beta-Elliptical Approach and Fuzzy Perceptual Detector

Parkinson’s Disease Detection From Online Handwriting Based on Beta-Elliptical Approach and Fuzzy Perceptual Detector

Parkinson's disease affects millions worldwide, and motor symptoms often appear years before clinical diagnosis — making early, non-invasive screening critically valuable. This project develops an intelligent diagnostic support system that analyzes handwriting dynamics captured from a digital tablet or stylus input. The system models handwriting strokes using the Beta-Elliptical approach, which represents pen velocity and trajectory through beta and Gaussian mathematical profiles, capturing micro-graphia, tremor, and kinematic irregularities characteristic of Parkinson's patients. A Fuzzy Perceptual Detector then classifies extracted kinematic features — speed, jerk, pen pressure, and stroke curvature — into healthy, at-risk, and Parkinson's-positive categories with interpretable confidence levels. Built using Python, OpenCV, and fuzzy logic toolkits with a web-based drawing interface, the system provides clinicians with instant screening scores and visual kinematic reports. This project stands out as a unique biomedical signal processing project, combining computational geometry, fuzzy inference, and machine learning — ideal for final-year students pursuing machine learning healthcare projects, neurodegenerative disease detection research, and AI-based diagnostic systems.

Components

Python 3.8+
OpenCV
Scikit-fuzzy (fuzzy logic toolkit)
NumPy & SciPy
Digital tablet / stylus input interface
Web-based drawing interface

Key Features

• Beta-Elliptical modeling of pen velocity and trajectory using beta and Gaussian profiles
• Capture of micro-graphia, tremor and kinematic irregularities from handwriting dynamics
• Fuzzy perceptual classification into healthy, at-risk and Parkinson's-positive categories
• Kinematic feature analysis — speed, jerk, pen pressure and stroke curvature
• Instant screening scores and visual kinematic reports for clinicians

Applications

Neurology clinics, motor-disorder researchers, biomedical signal processing students and health-tech startups building AI screening systems.