Hierarchical Contrastive Representation for Accurate Evaluation of Rehabilitation Exercises via Multi-View Skeletal Representations
Rehabilitation training is essential for the recovery of patients with conditions such as stroke and Parkinson’s disease. However, traditional skeletal-based assessments often fail to capture the subtle movement qualities necessary for personalized care and are not optimized for scoring t...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Transactions on Neural Systems and Rehabilitation Engineering |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10818452/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841557091922214912 |
|---|---|
| author | Zhejun Kuang Jingrui Wang Dawen Sun Jian Zhao Lijuan Shi Yusheng Zhu |
| author_facet | Zhejun Kuang Jingrui Wang Dawen Sun Jian Zhao Lijuan Shi Yusheng Zhu |
| author_sort | Zhejun Kuang |
| collection | DOAJ |
| description | Rehabilitation training is essential for the recovery of patients with conditions such as stroke and Parkinson’s disease. However, traditional skeletal-based assessments often fail to capture the subtle movement qualities necessary for personalized care and are not optimized for scoring tasks. To address these limitations, we propose a hierarchical contrastive learning framework that integrates multi-view skeletal data, combining both positional and angular joint information. This integration enhances the framework’s ability to detect subtle variations in movement during rehabilitation exercises. In addition, we introduce a novel contrastive loss function specifically designed for regression tasks. This new approach yields substantial improvements over existing state-of-the-art models, achieving over a 30% reduction in mean absolute deviation on both the KIMORE and UIPRMD datasets. The framework demonstrates robustness in capturing both global and local movement characteristics, which are critical for accurate clinical evaluations. By precisely quantifying action quality, the framework supports the development of more targeted, personalized rehabilitation plans and shows strong potential for broad application in rehabilitation practices as well as in a wider range of motion assessment tasks. |
| format | Article |
| id | doaj-art-86e143f7254d4ce59232592cf5e88b31 |
| institution | Kabale University |
| issn | 1534-4320 1558-0210 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Transactions on Neural Systems and Rehabilitation Engineering |
| spelling | doaj-art-86e143f7254d4ce59232592cf5e88b312025-01-07T00:00:14ZengIEEEIEEE Transactions on Neural Systems and Rehabilitation Engineering1534-43201558-02102025-01-013320121110.1109/TNSRE.2024.352390610818452Hierarchical Contrastive Representation for Accurate Evaluation of Rehabilitation Exercises via Multi-View Skeletal RepresentationsZhejun Kuang0https://orcid.org/0000-0001-5632-7596Jingrui Wang1https://orcid.org/0009-0006-2641-2099Dawen Sun2https://orcid.org/0009-0008-6309-2711Jian Zhao3https://orcid.org/0000-0003-3265-6461Lijuan Shi4Yusheng Zhu5College of Computer Science and Technology and the Key Laboratory of Intelligent Rehabilitation and Barrier-Free for the Disabled, Ministry of Education, Changchun University, Changchun, ChinaCollege of Computer Science and Technology and the Key Laboratory of Intelligent Rehabilitation and Barrier-Free for the Disabled, Ministry of Education, Changchun University, Changchun, ChinaCollege of Computer Science and Technology and the Key Laboratory of Intelligent Rehabilitation and Barrier-Free for the Disabled, Ministry of Education, Changchun University, Changchun, ChinaCollege of Computer Science and Technology and the Key Laboratory of Intelligent Rehabilitation and Barrier-Free for the Disabled, Ministry of Education, Changchun University, Changchun, ChinaJilin Provincial Key Laboratory of Human Health Status Identification Function and Enhancement, Changchun, ChinaCollege of Computer Science and Technology and the Key Laboratory of Intelligent Rehabilitation and Barrier-Free for the Disabled, Ministry of Education, Changchun University, Changchun, ChinaRehabilitation training is essential for the recovery of patients with conditions such as stroke and Parkinson’s disease. However, traditional skeletal-based assessments often fail to capture the subtle movement qualities necessary for personalized care and are not optimized for scoring tasks. To address these limitations, we propose a hierarchical contrastive learning framework that integrates multi-view skeletal data, combining both positional and angular joint information. This integration enhances the framework’s ability to detect subtle variations in movement during rehabilitation exercises. In addition, we introduce a novel contrastive loss function specifically designed for regression tasks. This new approach yields substantial improvements over existing state-of-the-art models, achieving over a 30% reduction in mean absolute deviation on both the KIMORE and UIPRMD datasets. The framework demonstrates robustness in capturing both global and local movement characteristics, which are critical for accurate clinical evaluations. By precisely quantifying action quality, the framework supports the development of more targeted, personalized rehabilitation plans and shows strong potential for broad application in rehabilitation practices as well as in a wider range of motion assessment tasks.https://ieeexplore.ieee.org/document/10818452/Physical rehabilitationaction quality assessmentcontrastive learningrepresentation learning |
| spellingShingle | Zhejun Kuang Jingrui Wang Dawen Sun Jian Zhao Lijuan Shi Yusheng Zhu Hierarchical Contrastive Representation for Accurate Evaluation of Rehabilitation Exercises via Multi-View Skeletal Representations IEEE Transactions on Neural Systems and Rehabilitation Engineering Physical rehabilitation action quality assessment contrastive learning representation learning |
| title | Hierarchical Contrastive Representation for Accurate Evaluation of Rehabilitation Exercises via Multi-View Skeletal Representations |
| title_full | Hierarchical Contrastive Representation for Accurate Evaluation of Rehabilitation Exercises via Multi-View Skeletal Representations |
| title_fullStr | Hierarchical Contrastive Representation for Accurate Evaluation of Rehabilitation Exercises via Multi-View Skeletal Representations |
| title_full_unstemmed | Hierarchical Contrastive Representation for Accurate Evaluation of Rehabilitation Exercises via Multi-View Skeletal Representations |
| title_short | Hierarchical Contrastive Representation for Accurate Evaluation of Rehabilitation Exercises via Multi-View Skeletal Representations |
| title_sort | hierarchical contrastive representation for accurate evaluation of rehabilitation exercises via multi view skeletal representations |
| topic | Physical rehabilitation action quality assessment contrastive learning representation learning |
| url | https://ieeexplore.ieee.org/document/10818452/ |
| work_keys_str_mv | AT zhejunkuang hierarchicalcontrastiverepresentationforaccurateevaluationofrehabilitationexercisesviamultiviewskeletalrepresentations AT jingruiwang hierarchicalcontrastiverepresentationforaccurateevaluationofrehabilitationexercisesviamultiviewskeletalrepresentations AT dawensun hierarchicalcontrastiverepresentationforaccurateevaluationofrehabilitationexercisesviamultiviewskeletalrepresentations AT jianzhao hierarchicalcontrastiverepresentationforaccurateevaluationofrehabilitationexercisesviamultiviewskeletalrepresentations AT lijuanshi hierarchicalcontrastiverepresentationforaccurateevaluationofrehabilitationexercisesviamultiviewskeletalrepresentations AT yushengzhu hierarchicalcontrastiverepresentationforaccurateevaluationofrehabilitationexercisesviamultiviewskeletalrepresentations |