The design and application of a genu valgum gait recognition model based on triple attention mechanism and spatial hierarchical pooling strategy_Journal of Biomedical Engineering

Authors：

SONG Xiaoneng1 ¹ ,  QIAN Kun2 ² , HOU Xuan3 ³ , WANG Yizhe ⁴

1. Department of Physical Education, Jiangnan University, Wuxi, Jiangsu 214122, P. R. China;
2. School of Artificial Intelligence and Computer Science, Jiangnan University, Wuxi, Jiangsu 214122, P. R. China;
3. School of Innovation Engineering, Macau University of Science and Technology, Macau SAR 0999078, P. R. China;
4. School of Science and Engineering, The Chinese University of Hong Kong, Shenzhen, Guangdong 518172, P. R. China;

Corresponding?author：

QIAN Kun2, Email: kqian@jiangnan.edu.cn

Keywords：

Genu valgum; Gait recognition; Triple attention module; Spatial hierarchical pooling module

DOI：

10.7507/1001-5515.202504005

Video：

Export PDF Favorites Scan Get Citation

Abstract Full text Figures/Tables Video References Cited by

To facilitate the early intelligent screening of pediatric genu valgum, this study develops a deep learning–based gait recognition model tailored for clinical application. The model is constructed upon a three-dimensional residual network architecture and incorporates a triplet attention module alongside a spatial hierarchical pooling module, jointly enhancing feature interaction across temporal, spatial, and channel dimensions. This design ensures an optimal balance between representational capacity and computational efficiency. Evaluated on a self-constructed dataset, the model achieves precision of 98.0%, 97.1%, and 96.5%, recall rates of 97.5%, 97.0%, and 95.0%, and F₁-scores of 0.980, 0.970, and 0.960 on the training, validation, and test sets, respectively, demonstrating excellent recognition performance and strong generalization ability. Ablation experiments confirm the importance of the proposed model’s core components in improving performance, and comparative experiments further highlight its significant advantages in recognition accuracy and robustness. Visualization experiments reveal that the model effectively focuses on key regions of gait images, with attention regions aligning closely with clinical anatomical landmarks, thereby enhancing the interpretability of the model’s decision-making in clinical applications. In summary, the proposed model not only offers an efficient and reliable technical solution for early intelligent screening of genu valgum in children, but also provides a practical pathway for applying gait recognition technology in medical diagnosis.

1.	World Health Organization. Working for a brighter, healthier future: how WHO improves health and promotes well-being for the world’s adolescents. Geneva, Switzerland, World Health Organization, 2024: 1-9.
2.	陶芳標. 構建面向健康中國2030青少年健康促進體系. 中國學校衛生, 2023, 44(1): 1-5.
3.	Dop D, P?dureanu V, P?dureanu R, et al. Risk factors involved in postural disorders in children and adolescents. Life, 2024, 14(11): 1463.
4.	魏銘, 牛雪松, 吳昊. 體醫融合視域下青少年體態異常防治的現實路徑. 沈陽體育學院學報, 2022, 41(4): 57-63.
5.	Rob C F, Hossain M D J, Kang X S. Management of post-traumatic genu valgus deformity in adolescent by proximal tibial corticotomy with bone lengthening by ilizarov method: a case report. MOJ Orthop Rheumatol, 2023, 15(6): 219-223.
6.	Buchan S, Bennet S, Barry M. Genu valgum in children. Orthopaedics and Trauma, 2022, 36(6): 311-316.
7.	Jamil K, Chew W Y, Bohari N E, et al. Knee measurements among children with normal alignment, physiologic and pathologic bowing aged 0-3 years old: a systematic review. Journal of Pediatric Orthopaedics B, 2022, 31(2): 105-113.
8.	Coppa V, Marinelli M, Procaccini R, et al. Coronal plane deformity around the knee in the skeletally immature population: a review of principles of evaluation and treatment. World Journal of Orthopedics, 2022, 13(5): 427-443.
9.	Moon S H, Kwon S S, Park M S, et al. Change of limb align-ment in Korean children and adolescents with idiopathic genu valgum. Medicine, 2021, 100(45): e27637.
10.	Di Gennaro G L, Trisolino G, Stallone S, et al. Guided growth technique for epiphysiodesis and hemiepiphysiodesis: safety and performance evaluation. Children, 2023, 11(1): 49.
11.	Sanchis G J B, Nascimento J A S D, Santana R C, et al. Biomechanical factors associated with patellofemoral pain in children and adolescents. Scientific Reports, 2024, 14(1): 15490.
12.	Khou S B, Saki F, Tahayori B. Muscle activation in the lower limb muscles in individuals with dynamic knee valgus during single-leg and overhead squats: a meta-analysis study. BMC Musculoskeletal Disorders, 2024, 25(1): 652.
13.	Lee S H, Choi Y, Lee J, et al. Valgus arthritic knee responds better to conservative treatment than the varus arthritic knee. Medicine, 2023, 59(4): 779.
14.	顧耀東, 徐異寧. 智能化兒童青少年體態健康評估預測模型開發——以小學齡兒童青少年膝關節內外翻為例. 上海體育大學學報, 2025, 49(4): 9-19.
15.	Ranade A S, Oka G A, Belthur M V, et al. An international consensus on evaluation and management of idiopathic genu valgum: a modified delphi survey. Journal of Pediatric Orthopaedics, 2025, 45(5): 274-280.
16.	房依婷, 黃雨琦, 于幸, 等. 雙任務范式在帕金森發病早期篩查中的應用進展. 中國療養醫學, 2023, 32(6): 591-594.
17.	Kirby J C, Jones H, Johnson B L, et al. Genu valgum in pediatric patients presenting with patellofemoral instability. Journal of Pediatric Orthopaedics, 2024, 44(3): 168-173.
18.	Meng D, He S, Wei M, et al. Enhanced predicting genu valgum through integrated feature extraction: utilizing ChatGPT with body landmarks. Biomedical Signal Processing and Control, 2024, 97: 106676.
19.	Munusamy V, Senthilkumar S. Emerging trends in gait recognition based on deep learning: a survey. Peer J Comput Sci, 2024, 10: e2158.
20.	Fan C, Hou S, Huang Y, et al. Exploring deep models for practical gait recognition. arXiv preprint, 2023, arXiv: 2303.03301.
21.	Franco A, Russo M, Amboni M, et al. The role of deep learning and gait analysis in Parkinson’s disease: a systematic review. Sensors, 2024, 24(18): 5957.
22.	Aghababa M P, Andrysek J. Exploration and demonstration of explainable machine learning models in prosthetic rehabilitation-based gait analysis. PloS One, 2024, 19(4): e0300447.
23.	Jeswani J, Shinde S, Singh K, et al. Innovative approaches in gait analysis for genu valgum using machine learning techniques. TechRxiv, 2024, DOI: 10.36227/techrxiv.173194435.53042495/v1.
24.	Bilal M, Jianbiao H, Mushtaq H, et al. GaitSTAR: spatial–temporal attention-based feature-reweighting architecture for human gait recognition. Mathematics, 2024, 12(16): 2458.
25.	Cui Y, Kang Y. Multi-modal gait recognition via effective spatial-temporal feature fusion//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Vancouver, Canada: IEEE, 2023: 17949-17957.
26.	Nie X, Chen X, Jin H, et al. Triplet attention transformer for spatiotemporal predictive learning//Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision. Hawaii: IEEE, 2024: 7036-7045.
27.	Hu Y, Yang J, Wang Y, et al. Multi-modal variable-channel spatial-temporal semantic action recognition network//International Conference of Pioneering Computer Scientists, Engineers and Educators (ICPCSEE 2024). Singapore: Springer Nature Singapore, 2024: 139-153.
28.	Junaid M I, Prakash A J, Ari S. Human gait recognition using joint spatiotemporal modulation in deep convolutional neural networks. Journal of Visual Communication and Image Representation, 2024, 105: 104322.
29.	Saoud L S, Hussain I. TempoNet: empowering long-term knee joint angle prediction with dynamic temporal attention in exoskeleton control//Proceedings of the 2023 IEEE-RAS 22nd International Conference on Humanoid Robots (Humanoids), Austin, USA: IEEE, 2023: 1-8.
30.	Ding G, Georgilas I, Plummer A. A deep learning model with a self-attention mechanism for leg joint angle estimation across varied locomotion modes. Sensors, 2023, 24(1): 211.
31.	Yan M, Guo M, Sun J, et al. Gait recognition in different terrains with IMUs based on attention mechanism feature fusion method. Neural Processing Letters, 2023, 55(8): 10215-10234.
32.	Wang M, Guo X, Lin B, et al. Dygait: exploiting dynamic representations for high-performance gait recognition//Proceedings of the IEEE/CVF International Conference on Computer Vision, Paris: IEEE, 2023: 13424-13433.
33.	Xu H, Zhang C, Wu Z, et al. PSGait: gait recognition using parsing skeleton. arXiv preprint, 2025, arXiv: 2503.12047.
34.	Salcedo E. Computer vision-based gait recognition on the edge: a survey on feature representations, models, and architectures. Journal of Imaging, 2024, 10(12): 326.
35.	Rajpurkar P, Chen E, Banerjee O, et al. AI in health and medicine. Nature Medicine, 2022, 28(1): 31-38.
36.	Ng S Y. Anonychia. The Journal of Pediatrics, 2023, 263: 113664.
37.	Yan J, Zheng B, Xu H, et al. Making pre-trained language models great on tabular prediction. arXiv preprint, 2024, arXiv: 2403.01841.
38.	An Y, Song C. Multiscale dynamic attention and hierarchical spatial aggregation for few-shot object detection. Applied Sciences, 2025, 15(3): 1381.
39.	Hong S B, Kim Y H, Nam S H, et al. S3D: squeeze and excitation 3D convolutional neural networks for a fall detection system. Mathematics, 2022, 10(3): 328.
40.	Woo S, Park J, Lee J, et al. CBAM: convolutional block attention module. arXiv preprint, 2018, arXiv: 1807.06521.
41.	Gao Z, Wang Q, Zhang B, et al. Temporal-attentive covariance pooling networks for video recognition. Advances in Neural Information Processing Systems, 2021, 34: 13587-13598.
42.	Sun Y, Long H, Feng X, et al. GaitASMS: gait recognition by adaptive structured spatial representation and multi-scale temporal aggregation. Neural Computing and Applications, 2024, 36(13): 7057-7069.
43.	Manoharan J, Sivagnanam Y. A novel human action recognition model by Grad-CAM visualization with multi-level feature extraction using global average pooling with sequence modeling by bidirectional gated recurrent units. International Journal of Computational Intelligence Systems, 2025, 18(1): 1-20.

1. World Health Organization. Working for a brighter, healthier future: how WHO improves health and promotes well-being for the world’s adolescents. Geneva, Switzerland, World Health Organization, 2024: 1-9.
2. 陶芳標. 構建面向健康中國2030青少年健康促進體系. 中國學校衛生, 2023, 44(1): 1-5.
3. Dop D, P?dureanu V, P?dureanu R, et al. Risk factors involved in postural disorders in children and adolescents. Life, 2024, 14(11): 1463.
4. 魏銘, 牛雪松, 吳昊. 體醫融合視域下青少年體態異常防治的現實路徑. 沈陽體育學院學報, 2022, 41(4): 57-63.
5. Rob C F, Hossain M D J, Kang X S. Management of post-traumatic genu valgus deformity in adolescent by proximal tibial corticotomy with bone lengthening by ilizarov method: a case report. MOJ Orthop Rheumatol, 2023, 15(6): 219-223.
6. Buchan S, Bennet S, Barry M. Genu valgum in children. Orthopaedics and Trauma, 2022, 36(6): 311-316.
7. Jamil K, Chew W Y, Bohari N E, et al. Knee measurements among children with normal alignment, physiologic and pathologic bowing aged 0-3 years old: a systematic review. Journal of Pediatric Orthopaedics B, 2022, 31(2): 105-113.
8. Coppa V, Marinelli M, Procaccini R, et al. Coronal plane deformity around the knee in the skeletally immature population: a review of principles of evaluation and treatment. World Journal of Orthopedics, 2022, 13(5): 427-443.
9. Moon S H, Kwon S S, Park M S, et al. Change of limb align-ment in Korean children and adolescents with idiopathic genu valgum. Medicine, 2021, 100(45): e27637.
10. Di Gennaro G L, Trisolino G, Stallone S, et al. Guided growth technique for epiphysiodesis and hemiepiphysiodesis: safety and performance evaluation. Children, 2023, 11(1): 49.
11. Sanchis G J B, Nascimento J A S D, Santana R C, et al. Biomechanical factors associated with patellofemoral pain in children and adolescents. Scientific Reports, 2024, 14(1): 15490.
12. Khou S B, Saki F, Tahayori B. Muscle activation in the lower limb muscles in individuals with dynamic knee valgus during single-leg and overhead squats: a meta-analysis study. BMC Musculoskeletal Disorders, 2024, 25(1): 652.
13. Lee S H, Choi Y, Lee J, et al. Valgus arthritic knee responds better to conservative treatment than the varus arthritic knee. Medicine, 2023, 59(4): 779.
14. 顧耀東, 徐異寧. 智能化兒童青少年體態健康評估預測模型開發——以小學齡兒童青少年膝關節內外翻為例. 上海體育大學學報, 2025, 49(4): 9-19.
15. Ranade A S, Oka G A, Belthur M V, et al. An international consensus on evaluation and management of idiopathic genu valgum: a modified delphi survey. Journal of Pediatric Orthopaedics, 2025, 45(5): 274-280.
16. 房依婷, 黃雨琦, 于幸, 等. 雙任務范式在帕金森發病早期篩查中的應用進展. 中國療養醫學, 2023, 32(6): 591-594.
17. Kirby J C, Jones H, Johnson B L, et al. Genu valgum in pediatric patients presenting with patellofemoral instability. Journal of Pediatric Orthopaedics, 2024, 44(3): 168-173.
18. Meng D, He S, Wei M, et al. Enhanced predicting genu valgum through integrated feature extraction: utilizing ChatGPT with body landmarks. Biomedical Signal Processing and Control, 2024, 97: 106676.
19. Munusamy V, Senthilkumar S. Emerging trends in gait recognition based on deep learning: a survey. Peer J Comput Sci, 2024, 10: e2158.
20. Fan C, Hou S, Huang Y, et al. Exploring deep models for practical gait recognition. arXiv preprint, 2023, arXiv: 2303.03301.
21. Franco A, Russo M, Amboni M, et al. The role of deep learning and gait analysis in Parkinson’s disease: a systematic review. Sensors, 2024, 24(18): 5957.
22. Aghababa M P, Andrysek J. Exploration and demonstration of explainable machine learning models in prosthetic rehabilitation-based gait analysis. PloS One, 2024, 19(4): e0300447.
23. Jeswani J, Shinde S, Singh K, et al. Innovative approaches in gait analysis for genu valgum using machine learning techniques. TechRxiv, 2024, DOI: 10.36227/techrxiv.173194435.53042495/v1.
24. Bilal M, Jianbiao H, Mushtaq H, et al. GaitSTAR: spatial–temporal attention-based feature-reweighting architecture for human gait recognition. Mathematics, 2024, 12(16): 2458.
25. Cui Y, Kang Y. Multi-modal gait recognition via effective spatial-temporal feature fusion//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Vancouver, Canada: IEEE, 2023: 17949-17957.
26. Nie X, Chen X, Jin H, et al. Triplet attention transformer for spatiotemporal predictive learning//Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision. Hawaii: IEEE, 2024: 7036-7045.
27. Hu Y, Yang J, Wang Y, et al. Multi-modal variable-channel spatial-temporal semantic action recognition network//International Conference of Pioneering Computer Scientists, Engineers and Educators (ICPCSEE 2024). Singapore: Springer Nature Singapore, 2024: 139-153.
28. Junaid M I, Prakash A J, Ari S. Human gait recognition using joint spatiotemporal modulation in deep convolutional neural networks. Journal of Visual Communication and Image Representation, 2024, 105: 104322.
29. Saoud L S, Hussain I. TempoNet: empowering long-term knee joint angle prediction with dynamic temporal attention in exoskeleton control//Proceedings of the 2023 IEEE-RAS 22nd International Conference on Humanoid Robots (Humanoids), Austin, USA: IEEE, 2023: 1-8.
30. Ding G, Georgilas I, Plummer A. A deep learning model with a self-attention mechanism for leg joint angle estimation across varied locomotion modes. Sensors, 2023, 24(1): 211.
31. Yan M, Guo M, Sun J, et al. Gait recognition in different terrains with IMUs based on attention mechanism feature fusion method. Neural Processing Letters, 2023, 55(8): 10215-10234.
32. Wang M, Guo X, Lin B, et al. Dygait: exploiting dynamic representations for high-performance gait recognition//Proceedings of the IEEE/CVF International Conference on Computer Vision, Paris: IEEE, 2023: 13424-13433.
33. Xu H, Zhang C, Wu Z, et al. PSGait: gait recognition using parsing skeleton. arXiv preprint, 2025, arXiv: 2503.12047.
34. Salcedo E. Computer vision-based gait recognition on the edge: a survey on feature representations, models, and architectures. Journal of Imaging, 2024, 10(12): 326.
35. Rajpurkar P, Chen E, Banerjee O, et al. AI in health and medicine. Nature Medicine, 2022, 28(1): 31-38.
36. Ng S Y. Anonychia. The Journal of Pediatrics, 2023, 263: 113664.
37. Yan J, Zheng B, Xu H, et al. Making pre-trained language models great on tabular prediction. arXiv preprint, 2024, arXiv: 2403.01841.
38. An Y, Song C. Multiscale dynamic attention and hierarchical spatial aggregation for few-shot object detection. Applied Sciences, 2025, 15(3): 1381.
39. Hong S B, Kim Y H, Nam S H, et al. S3D: squeeze and excitation 3D convolutional neural networks for a fall detection system. Mathematics, 2022, 10(3): 328.
40. Woo S, Park J, Lee J, et al. CBAM: convolutional block attention module. arXiv preprint, 2018, arXiv: 1807.06521.
41. Gao Z, Wang Q, Zhang B, et al. Temporal-attentive covariance pooling networks for video recognition. Advances in Neural Information Processing Systems, 2021, 34: 13587-13598.
42. Sun Y, Long H, Feng X, et al. GaitASMS: gait recognition by adaptive structured spatial representation and multi-scale temporal aggregation. Neural Computing and Applications, 2024, 36(13): 7057-7069.
43. Manoharan J, Sivagnanam Y. A novel human action recognition model by Grad-CAM visualization with multi-level feature extraction using global average pooling with sequence modeling by bidirectional gated recurrent units. International Journal of Computational Intelligence Systems, 2025, 18(1): 1-20.

Journal of Biomedical Engineering

Latest ArticlesThe design and application of a genu valgum gait recognition model based on triple attention mechanism and spatial hierarchical pooling strategy

Abstract Full text Figures/Tables Video References Cited by

Format

Content