Diagnosing Parkinson's disease via behavioral biometrics of keystroke dynamics.

Abstract

Parkinson's disease (PD) is one of the rapidly growing neurodegenerative diseases, affecting more than 10 million people worldwide. Early and accurate diagnosis of PD is highly desirable for therapeutic interventions but remains a substantial challenge. We developed a soft, portable intelligent keyboard leveraging magnetoelasticity to detect subtle pressure variations in keystroke dynamics by converting continuous keystrokes into high-fidelity electrical signals, thus enabling the quantitative analysis of PD motor symptoms using machine learning. Relying on a fundamental working mechanism, the intelligent keyboard demonstrates highly sensitive, intrinsically waterproof, and biocompatible properties, with the successful demonstration in a pilot study on patients with PD. To facilitate the potential continuous monitoring of PD, a customized cellphone application was developed to integrate the intelligent keyboard into a wireless platform. Together, the intelligent keyboard system's compelling properties position it as a promising tool for advancing early diagnosis and facilitating personalized, predictive, preventative, and participatory approaches to PD healthcare.