Vision Optical-Based Evaluation of Senshand Accuracy for Parkinson’s Disease Motor Assessment

Abstract

Parkinson’s disease (PD) is a highly-disabling common neurodegenerative disease. The current method to diagnose PD is mainly based on clinical criteria and motor examination of the performance of the patients that are visually evaluated by the neurologist while they performed tasks from MDS-UPDRS. In order to support the clinician in objective assessment of the motor performance, sensorized wearable technology able to finely measure the motion are currently investigated. Since accuracy and precision of the measures are mandatory to provide the neurologist with a tool that can actually be applied in clinical practice, in this work, the quality of measures obtained by the SensHand system was evaluated, comparing them to an optoelectronic “gold standard” system addressing a preliminary technical validation. Three exercises (i.e., finger tapping, thumb-forefinger tapping, and pronosupination) were selected and frequency, number of repetitions and amplitude of the movements were measured for each of them by both the wearable and optical systems. The preliminary results were very satisfying, considering that discrepancies, measured as absolute error, in frequency evaluation, number of repetitions, and amplitude movement were, on average, equal to 0.03 taps/s(min0.02, max0.05), 0.07 tap (min 0.02, max 0.13), and 3.8 degrees (min 1.81, max7.47), respectively. Very high correlation values obtained from linear regression analysis (R2>0.9) also confirmed the accuracy of the measurements achieved with SensHand. Therefore, the obtained results from SensHand system are promising to use it to support the neurologist for accurate quantification of motion analysis to improve the objective PD evaluation.