Previously submitted to: JMIR Medical Informatics (no longer under consideration since Jan 10, 2020)
Date Submitted: Oct 18, 2019
Warning: This is an author submission that is not peer-reviewed or edited. Preprints - unless they show as "accepted" - should not be relied on to guide clinical practice or health-related behavior and should not be reported in news media as established information.
A system for automated motor function assessment in stroke patients using motion sensing technology
ABSTRACT
Background:
Upper-limb motor function assessment for stroke patients is essential for planning interventions to maximise function and independence.
Objective:
The general aim of this study is to realize the automated assessment of upper-limb motor function after stroke by using motion sensing technology and machine learning algorithms.
Methods:
The proposed system contained two subsystems. A motion tracking subsystem was developed for measuring the kinematic data of participants through Kinect V2. A motor function assessment subsystem was developed to realize the automated assessment based on a feed-forward neural network (FFNN)-based assessment model. The assessment model was developed by thirty-two kinematic metrics calculated from the gathered data. In addition, five metrics with ten-fold cross-validation were used for performance evaluation (accuracy, F1-score, specificity, sensitivity, and AUC), and the developed FFNN-based model was compared with ten well-known machine learning algorithm-based models.
Results:
26 participants between 20 to 79 years of age, living in the Fuzhou city (Fujian) were recruited from October 2018 to January 2019. A total of 2400 patterns were calculated as the dataset after data processing. The experimental results showed that the proposed system presented satisfactory performance: overall accuracy ranged from 0.87 to 0.96, and F1-score ranged from 0.83 to 0.93. The FFNN-based assessment model presented promising comprehensive performance: it ranked top two for all tasks in the metrics of overall accuracy and F1-score.
Conclusions:
The proposed system presents a promising method for automated assessment of upper-limb motor function in stroke patients. The system could be implemented in clinical applications as a cheap and portable assessment tool and provide decision support for physiotherapists as well.
Citation
Request queued. Please wait while the file is being generated. It may take some time.
Copyright
© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.