We use cookies to ensure that we give you the best experience on our website Learn more

Home

Saved research

Submission

AUTOMATIC IDENTIFICATION OF MOTOR ANOMALIES IN STROKE PATIENTS WITH CUSTOMIZED COMPUTATIONAL MOVEMENT SYSTEM.

Submitted

and 2 other(s)

4 Views
0 Downloads
0 Saves

Presented at

ESOC-2019

-

Presentation

thumbnail

Abstract

Background and Aims: Motion Capture Systems (MCS) are currently use in stroke rehabilitation, but its use in stroke diagnosis is still undeveloped. Our aim is to test a customized MCS’ algorithm that automatically identify motor deficit in stroke patients. Methods: Using Kinect camera (Microsoft) and customized software Akira (system friend Inc.) we registered 10 exercises divided in 3 workouts: the first one included trunk stability and walk, the second one included right upper limb exercises and the third one includes left upper limb exercises. We designed a case-control study, in which compared the performance of the exercises between stroke patients (cases) and healthy controls. The controls performance was registered and obtained normality movements values. Then, with a customized comparison algorithm (developed with MATLAB), we automatically compared the cases movement with controls and obtained a complete report of movement trajectories and their deviation. Results: We analyzed 30 healthy controls and 14 stroke patients (median NIHSS 2, IQR 0-12): 6 with left hemisphere damage, 6 with right hemisphere damage, 1 with bilateral damage and 1 with cerebellar damage. All stroke patients had alterations in trunk stability (first workout): right hemisphere damage patients tend to imbalance forward, while left hemisphere damage patients tend to imbalance to the left. In the upper limb evaluation all stroke patients had right or left alteration according to brain damage, except in 2 right damage patients that showed alteration in both arms. Conclusions: Our system was useful in detecting movement alterations in stroke patients with higher precision than clinical exploration

Datasets

No datasets are available for this submission.

License

No license information is available for this submission.

Morressier

Company

Legal

Follow us

© Copyright 2020 Morressier GmbH. All rights reserved.

Morressier

© Copyright 2020 Morressier GmbH.
All rights reserved.