The Clinical Neurologist International 2020 | Volume 2 | Article 1009 05 © 2020 - Medtext Publications. All Rights Reserved. ISSN: 2691-5669 Automated Assessment of Motor Impairments in Parkinson’s Disease Research Article Claudia Ferraris 1 , Roberto Nerino 1 , Antonio Chimienti 1 , Giuseppe Pettiti 1 , Corrado Azzaro 2 , Giovanni Albani 2 , Alessandro Mauro 2,3 and Lorenzo Priano 2,3,* 1 Deartment of Computer and Telecommunication Engineering, Institute of Electronics, National Research Council, Torino, Italy 2 Department of Neurology and Neurorehabilitation, Istituto Auxologico Italiano, IRCCS, S.Giuseppe Hospital, Piancavallo, (VB), Italy 3 Department of Neurosciences, University of Turin, Torino, Italy Citation: Ferraris C, Nerino R, Chimienti A, Pettiti G, Azzaro C, Albani G, et al. Automated Assessment of Motor Impairments in Parkinson’s Disease. Clin Neurol Int. 2020; 2(1): 1010. Copyright: © 2020 Claudia Ferraris Publisher Name: Medtext Publications LLC Manuscript compiled: Feb 14 th , 2020 *Corresponding author: Lorenzo Priano, Department of Neurology and Neurorehabilitation, Istituto Auxologico Italiano, IRCCS, S.Giuseppe Hospital, Piancavallo, (VB), Italy, E-mail: lorenzo.priano@unito.it Abstract A system for the automatic assessment of motor impairments in Parkinson’s Disease (PD) is presented. Te interface, built around optical RGB-Depth devices, allows for tracking of hands and body movements during the performance of standard upper and lower limb tasks, as specifed by the Unifed Parkinson’s Disease Rating Scale (UPDRS). Te assessment of the diferent tasks is performed by machine learning techniques. Selected kinematic parameters characterizing the movements are input to trained classifers to rate the motor performance. Te accurate tracking and characterization of the movements allows for an automatic and objective assessment of the UPDRS tasks, making feasible the monitoring of motor fuctuations also at-home for telemedicine or neurorehabilitation purposes. Keywords: Parkinson’s disease; Movement disorders; UPDRS; Automated assessment; Natural human computer interface; RGB-Depth; At-home monitoring; Hand tracking; Body tracking Introduction Parkinson's Disease (PD) is a neurodegenerative disease characterized by a progressive motor impairments, whose severity is subjectively assessed by clinicians during the performance of standard motor tasks usually defned by the motor examination section of the Unifed Parkinson's Disease Rating Scale (UPDRS) [1,2]. Objective and automatic assessment of the tasks at-home can improve the reliability of the assessment, generally infuenced by inter-rater disagreements [3], and could allow a weekly adjustment of the therapy, reducing fuctuations [4,5]. Proposed solutions for the automatic assessment of PD motor tasks make use of the established correlation existing between kinematic characteristics of the movements and the severity of the impairment [4,5], mainly by technologies based on optical devices and wearable sensors [6,7]. Approaches based on wearable sensors require the involvement of the patient for the initial setup and, possibly, for the calibration phase, are usually uncomfortable for impaired people and their invasiveness can afect functional performance [8,9]. On the contrary, optical approaches based on recent RGB-Depth devices are less invasive and allow for accurate measurements, so they have been proposed for tracking the body and hand movements in the framework of PD assessment [10,11]. In this context, we present a low-cost system for the automatic and at-home assessment of some of the upper and lower limb tasks of the UPDRS, namely Finger Tapping (FT), Opening-Closing (OC) of the hand and Pronation-Supination (PS) of the hand, Sit-to-Stand (S2S) and Leg Agility (LA). Te system implements a non-invasive gesture- based Human Computer Interface (HCI), which allows people with motor impairments both to interact with the graphical interface of the system through simple gestures such as opening and closing of the hand or pointing with fngers on interactive objects, and the tracking of hands and body movements, for the assessment of the motor performance during the performance of standard UPDRS tasks. In particular, the developed algorithm for the hand tracking has proved to be more robust and accurate for fast movements respect to other solutions based on proprietary algorithms provided by commercial devices [10], making the assessment more reliable. In addition, the algorithm for the hand tracking does not depend on any particular device or proprietary Sofware Development Kit (SDK), but requires only the RGB and depth information availability at a proper frame rate. Results on experiments performed to validate the system are presented: the accuracies obtained in the automatic assessment of the considered UPDRS tasks, as compared to the clinician standard assessment, demonstrate the feasibility of the system also for the at- home remote monitoring of PD motor impairments. Patients and Methods Two cohorts of 44 PD patients (mean Hoehn and Yahr score 2.3, min 1, max 4; age 41-85 years; disease duration 1-29 years), and 15 Healthy Control (HC) subjects respectively (age 45-78 years) were recruited. Patients were excluded if they had tremor severity >1 or cognitive impairment (Mini-Mental State Examination Score <27/30). All subjects provided their informed consent prior to their participation. Te PD cohort was assessed by two neurologists with experience in movement disorders. Te system hardware consists of two diferent setups: a near mode operation setup for the capture and the assessment of upper limb tasks, and a far mode setup for the capture and the assessment of lower limb tasks.