This paper demonstrated that it is possible to distinguish severity of bradykinesia using muscle vibration information. Muscle vibration information was captured in mechanomyography (MMG) signal using wearable accelerometer sensors. Two groups of 4 Parkinson’s subjects, less and more severe groups, had participated in this study. Each participant performed 4 tasks: finger pinching, hand opening and closing, hand pronation and supination, and wrist flexion and extension. PDPack toolbox was used for event detection and signal processing. Hilbert spectrum was estimated from the resultant signal from three axes of the accelerometer to obtained the power spectrum. Instantaneous mean frequency (IMNF) was then calculated for comparison. The results indicate that more severe bradykinesia has lower IMNF. The difference between the two groups is also visible from the Hilbert spectrum plots.

XII SIMPÓSIO DE ENGENHARIA BIOMÉDICA - IX SIMPÓSIO DE INSTRUMENTAÇÃO E IMAGENS MÉDICAS