Exploring Data Analysis To Assess The Effects Of Telerehabilitation For People With Parkinson’s Disease

Parkinson’s disease (PD) is a common neurodegenerative and progressive disorder affecting millions of people world wild, that was firstly described in 1817 by Dr. James Parkinson. Even though more than 200 years have passed from the PD discovery, an effective cure, able to slow down, stop or modify the disease course, has not been found yet, meaning that all the pharmacological medications available nowadays are destined to alleviate the main PD motor and nonmotor symptoms. Currently, physical activity and exercise are achieving greater relevance as promising and effective alternative treatments to complement the prescribed drug therapy, thanks to many demonstrated benefits that they could induce in those affected. However, due to the many restrictions imposed by the recent COVID-19 emergency, PD patients could not get any more access to fundamental visits and cares, including physical rehabilitation. To respond to the changed healthcare needs imposed by the pandemic, RAPIDO (“teleReabilitatione per I mAlati di Parkinson In qualsiasi staDIO”) has proposed and developed a telerehabilitation and telemonitoring systems destinated to people suffering from PD, relying on easy-to-access and economically-sustainable devices. In particular, a common Samsung A7 tablet delivers the rehabilitation services, while a Garmin smartwatch continuously collect daily and nocturnal patient-related data. Basing on the gathered information about the enrolled subjects, this experimental dissertation aims to define a data analysis method to assess whether the telerehabilitation protocol, proposed by RAPIDO, has induced some changes in the state of health of enrolled PD patients, supporting the scale-bases clinical evaluations effectuated by medical doctors. Specifically, the presented analysis approach involves, firstly, a statistical analysis to identify the variables that significantly change throughout the telerehabilitation period and, secondly, an unsupervised clustering technique that could provide further insights about the progression of the disease during the RAPIDO program. The obtained results suggest that the proposed method has not been able to depicts any variation of the processed data during the telerehabilitation period for any of the considered patients. Nonetheless, these outcomes seem to be in line with subjects’ actual medical conditions; indeed, the evaluated clinical parameters indicate that the patient’s states of health have remained stable throughout the entire at-home rehabilitation program, despite the strong degenerative nature of the disease itself. In conclusion, the training services proposed by RAPIDO, in combination with an effective pharmacological treatment, seems to contribute the slowdown of the PD course even in patients characterized by different initial disease stages.