Hanny, L.S.A. (Stefan) (2020) Game-based training for pattern recognition control based prosthesis: A comparison between conventional and game-based training on performance in daily-life tasks with a pattern recognition based prosthesis. thesis, Human Movement Sciences.
Full text available on request.Abstract
Objective: Pattern recognition controlled prostheses require prolonged training and practice before they can be properly controlled. The goal of this training is to improve distinguishability between movement patterns. This distinguishability between patterns is required, because the controller has to be able to discern the different patterns. This distinguishability is achieved by continuous repetitions of the same movements, making the training rather dull after a while. Due to the long training process, motivation loss is one of the main shortcomings of current training programs. Serious games are suggested as an alternative training method, to increase motivation and provide feedback during training. The aim of this study was to explore whether game-based training would increase performance in daily tasks with pattern recognition controlled prosthesis and compare this to conventional training. Methods: Six participants participated in 10 training sessions over the course of one month. Half of the participants received game-based training and the other received half conventional training . Prosthesis performance was tested using the SHAP and clothespin tests twice, during first and last training session. To determine the distinguishability of muscle activation patterns, interclass distance of electromyography (EMG) patterns in feature space was calculated with the separability index. Results: No difference in functional performance during the SHAP and clothespin relocation tests was found in either groups from pre- to post-test. Furthermore, no significant difference was found between the training groups. The interclass distance of EMG patterns did not change significantly either. Conclusion: Game-based and conventional training of myoelectric prosthetics did not yield significantly increased performance on either functional test or on separability of the patterns. However, participants reported feeling more comfortable and skillful with the device after training. Significance: Insight in game-based training for pattern recognition prosthesis is important to find opportunities to make the user training more enjoyable for the participant. In addition, the option to provide the user with individualized feedback Index Terms – clothespin relocation test, conventional training, game-based training, pattern recognition, prosthesis, separability index, serious games, SHAP test.
| Item Type: | Thesis (UNSPECIFIED) |
|---|---|
| Faculty: | Medical Sciences |
| Date Deposited: | 18 May 2022 09:46 |
| Last Modified: | 18 May 2022 09:46 |
| URI: | https://umcg.studenttheses.ub.rug.nl/id/eprint/3350 |
Actions (login required)
 |
View Item |