DEVELOPMENT OF REMO SOLUTION FOR THE RECOVERY OF HAND MOTOR FUNCTION AFTER STROKE, IN REMOTE REHABILITATION SETTINGS

G. Pregnolato¹, L. Privitera², S. Giordano², G. Nicoletti Rosa², P. Ariano², A. Turolla^3,4

¹IRCCS San Camillo Hospital, Laboratory of Rehabilitation Technologies, Venice, Italy, ²Morecognition Ltd., Turin, Italy, ³Alma Mater Studiorum Università di Bologna, Department of Biomedical and Neuromotor Sciences – DIBINEM, Bologna, Italy, ⁴IRCCS Policlinico Sant’Orsola-Malpighi, Division of Occupational Medicine, Bologna, Italy

Background: In the last decade, technology-based solutions have been growing up for application in rehabilitation after stroke. Specifically, the use of sensors on subjects’ body parts should provide data to physical therapist on patient’s motor performance. Moreover, electromyographic sensors provide feedback to patients on their muscle activity while moving. In rehabilitation field, smartphone App solutions are used to control wearable device aimed to enhance physical activity in elderly people, or in subjects affected by neurological disease. In this study, we developed the App REMO for Android, allowing physical therapists to plan rehabilitation training for patients and monitoring their motor performance.

Purpose: The aim of the study is to test the motor performance indexes for monitoring hand motor training by an innovative wearable device, in a remote setting.

Methods: Remo (Morecognition Ltd., Turin, Italy) is a wearable device able to detect forearm muscles activation (i.e. surface electromyography). By exploiting the Non-Negative Matrix Factorization (NMF), a new algorithm is used to extract a measure of the quality of movement, which would provide the percentage of similarity between muscle activation in real time and the best movement of the patient. In order to test the algorithm, we asked to 10 right-handed healthy subjects to perform 21 hand movements for 5 repetitions each. The movements are the following: 6 wrist movements in two different shoulder and elbow postures (i.e. shoulder 0° and elbow flexion 90°; shoulder flexion 30° and elbow 0°), 4 fingers task without any object to manipulate and 5 different grasps. We used the NMF algorithm to test the similarity between 21 gestures. We imposed three level of similarity: high (from 100% to 70%), moderate (from 70% to 30%) and low (from 30% to 0%).

Results: After sEMG acquisition, data showed that there is a high level of similarity between the same movements performed in different position (wrist extension 85.8%; wrist flexion 74.6%) and between two type of grasps (grasping a paper sheet; grasping a pen: 78.6%). Furthermore, there is a moderate level of similarity between all the type of grasping and fingers flexion (from 64.3% and 49.1%). Finally, there is a low similarity between wrist flexion and grasps (from 9.6% and 15.5%).

Conclusions: REMO is able to classify different hand movements and to test the similarity between wrist and hand movements. Future developments includes testing, in a randomised-control trial, efficacy of using REMO as a device for rehabilitation of hand motor recovery, after stroke.

Implications: REMO device is an innovative solution to promote hand rehabilitation in people with a stroke. The App solution provide meaningful data to physical therapists for monitoring patients’ muscle activation in remote settings included telerehabilitation applications.

Funding acknowledgements: The work is supported by Morecognition Ltd (Turin, Italy).

Keywords:
Wearable device
Surface Electromyography
Neurorehabilitation

Topics:
Innovative technology: robotics
Innovative technology: information management, big data and artificial intelligence
Neurology: stroke

Did this work require ethics approval? Yes

Institution: IRCCS San Camillo Hospital, Venice, Italy

Committee: Ethics Committee of Venice and IRCCS San Camillo Hospital

Ethics number: CE protocol no. 2021.13