Rehabilitation Exergames: Use of Motion Sensing and Machine Learning to Quantify Exercise Performance in Healthy...

Area of Science:

• Biomedical Engineering
• Rehabilitation Technology
• Machine Learning in Healthcare

Background:

• Home physiotherapy lacks objective performance feedback, potentially leading to incorrect exercise execution and worsened conditions.
• Current qualitative assessments by physiotherapists are not available during home-based rehabilitation.
• Objective performance tracking is crucial for effective patient self-monitoring and clinical progress assessment.

Purpose of the Study:

• To propose and evaluate machine learning algorithms for quantitative assessment of physiotherapy exercise performance.
• To compare the efficacy of dynamic time warping (DTW) and hidden Markov model (HMM) in objectively measuring patient movements against a reference.
• To enable progress tracking for both patients and clinicians in a home-based physiotherapy setting.

Main Methods:

• Utilized a Kinect V2 motion sensor to capture 3D skeletal movement data of 16 participants performing 4 exercises.
• Compared patient movement data against a reference standard (physiotherapist performance).
• Applied dynamic time warping (DTW) and hidden Markov model (HMM) algorithms to analyze movement data and generate performance scores.

Main Results:

• Both DTW and HMM algorithms demonstrated a similar trend in assessing participant performance.
• DTW showed higher sensitivity to minor movement variations, while HMM provided a more general overview of performance.
• The algorithms successfully generated quantitative scores reflecting exercise execution quality.

Conclusions:

• Machine learning algorithms (HMM and DTW) can objectively assess physiotherapy exercise performance.
• HMM is suitable for early-stage general performance evaluation, while DTW excels at detailed analysis later in therapy.
• Generated performance scores facilitate patient self-monitoring and provide valuable data for physiotherapists to track progress and adjust treatment plans.