基于时变协同的脑瘫患儿肌电包络信号生成模型研究
Research on a Time-Varying Synergy-Based Model for sEMG Envelope Signal Generation
摘要: 表面肌电信号(surface electromyography, sEMG)是由肌肉活动引发的微弱电信号,由于sEMG信号采集过程受试者募集难度及伦理审批的限制,快速获取大规模样本用于研究面临挑战。本文为解决脑瘫(Cerebral Palsy, CP)患者样本不足的问题,提出了一种以时变肌肉协同模型为基础的CP患儿肌电包络信号生成的模型。将采集的健康成人和CP患儿步态过程双下肢各8块肌肉sEMG信号和加速度信号,通过非负矩阵算法提取受试者的协同结构矩阵和激活系数矩阵,并以健康成人的激活系数矩阵为参考,通过引入时间延迟变量获得一系列更新后的激活系数矩阵,随后与CP患儿的协同结构矩阵相乘,进而得到全新的肌电包络信号。为验证生成信号的准确性,本文通过交并比(Intersection over Union, IoU) IoU评估生成信号的肌肉激活时间与真实CP患者之间的相似性,结果显示8块肌肉的平均相似性为64.94%,表明生成包络信号符合CP患者的肌肉激活特征。
Abstract: Surface electromyography (sEMG) signals, which are weak electrical signals generated by muscle activity, face challenges in rapid acquisition of large-scale samples for research due to difficulties in participant recruitment and ethical approval restrictions during the signal acquisition process. To address the issue of insufficient samples from cerebral palsy (CP) patients, this paper proposes a model for generating sEMG envelope signals for children with CP based on a time-varying muscle synergy model. sEMG and acceleration signals from eight muscles in both lower limbs during gait were collected from healthy adults and children with CP. The synergy structure matrix and activation coefficient matrix of the subjects were extracted using a non-negative matrix factorization algorithm. Using the activation coefficient matrix of healthy adults as a reference, a series of updated activation coefficient matrices were obtained by introducing time delay variables. These matrices were then multiplied with the synergy structure matrix of children with CP to generate new sEMG envelope signals. To validate the accuracy of the generated signals, the Intersection over Union (IoU) was used to evaluate the similarity between the muscle activation timing of the generated signals and that of real CP patients. The results showed an average similarity of 64.94% across the eight muscles, demonstrating that the generated envelope signals align well with the muscle activation characteristics of CP patients.
文章引用:汤璐, 郑辉, 王祥瑞, 潘子豪. 基于时变协同的脑瘫患儿肌电包络信号生成模型研究[J]. 建模与仿真, 2025, 14(3): 135-145. https://doi.org/10.12677/mos.2025.143209

参考文献

[1] 陈秀洁, 李树春. 小儿脑性瘫痪的定义、分型和诊断条件[J]. 中华物理医学与康复杂志, 2007, 29(5): 309.
[2] Patel, D.R., Neelakantan, M., Pandher, K. and Merrick, J. (2020) Cerebral Palsy in Children: A Clinical Overview. Translational Pediatrics, 9, S125-S135. [Google Scholar] [CrossRef] [PubMed]
[3] Schmidt‐Rohlfing, B., Bergamo, F., Williams, S., Erli, H.J., Rau, G., Niethard, F.U., et al. (2006) Interpretation of Surface EMGs in Children with Cerebral Palsy: An Initial Study Using a Fuzzy Expert System. Journal of Orthopaedic Research, 24, 438-447. [Google Scholar] [CrossRef] [PubMed]
[4] Michelsen, J.S., Lund, M.C., Alkjær, T., Finni, T., Nielsen, J.B. and Lorentzen, J. (2020) Wearable Electromyography Recordings during Daily Life Activities in Children with Cerebral Palsy. Developmental Medicine & Child Neurology, 62, 714-722. [Google Scholar] [CrossRef] [PubMed]
[5] Farina, D. and Merletti, R. (2001) A Novel Approach for Precise Simulation of the EMG Signal Detected by Surface Electrodes. IEEE Transactions on Biomedical Engineering, 48, 637-646. [Google Scholar] [CrossRef] [PubMed]
[6] Farina, D., Mesin, L., Martina, S. and Merletti, R. (2004) A Surface EMG Generation Model with Multilayer Cylindrical Description of the Volume Conductor. IEEE Transactions on Biomedical Engineering, 51, 415-426. [Google Scholar] [CrossRef] [PubMed]
[7] Campbell, E., Cameron, J.A.D. and Scheme, E. (2020). Feasibility of Data-Driven EMG Signal Generation Using a Deep Generative Model. 2020 42nd Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC), Montreal, 20-24 July 2020, 3755-3758.[CrossRef] [PubMed]
[8] Zhu, F., Ye, F., Fu, Y., Liu, Q. and Shen, B. (2019) Electrocardiogram Generation with a Bidirectional LSTM-CNN Generative Adversarial Network. Scientific Reports, 9, Article No. 6734. [Google Scholar] [CrossRef] [PubMed]
[9] Hazra, D. and Byun, Y. (2020) SynSigGAN: Generative Adversarial Networks for Synthetic Biomedical Signal Generation. Biology, 9, Article No. 441. [Google Scholar] [CrossRef] [PubMed]
[10] 钱尧佳. 基于对抗神经网络的肌电信号生成模型研究[D]: [硕士学位论文]. 杭州: 电子科技大学, 2024.
[11] 董梦坤. sEMG的特征分析与肌肉疲劳分类研究[D]: [硕士学位论文]. 长春: 吉林大学, 2022.
[12] Campanini, I., Merlo, A., Degola, P., Merletti, R., Vezzosi, G. and Farina, D. (2007) Effect of Electrode Location on EMG Signal Envelope in Leg Muscles during Gait. Journal of Electromyography and Kinesiology, 17, 515-526. [Google Scholar] [CrossRef] [PubMed]
[13] Di Nardo, F. and Fioretti, S. (2024) Accuracy of EMG Linear Envelope in Identifying the Peak of Muscular Activity during Walking. Gait & Posture, 111, 185-190. [Google Scholar] [CrossRef] [PubMed]
[14] 李飞. 基于表面肌电信号的小儿脑瘫步态肌肉协同分析[D]: [硕士学位论文]. 合肥: 中国科学技术大学, 2014.
[15] D’Avella, A., Saltiel, P. and Bizzi, E. (2003) Combinations of Muscle Synergies in the Construction of a Natural Motor Behavior. Nature Neuroscience, 6, 300-308. [Google Scholar] [CrossRef] [PubMed]
[16] Fox, E.J., Tester, N.J., Kautz, S.A., Howland, D.R., Clark, D.J., Garvan, C., et al. (2013) Modular Control of Varied Locomotor Tasks in Children with Incomplete Spinal Cord Injuries. Journal of Neurophysiology, 110, 1415-1425. [Google Scholar] [CrossRef] [PubMed]
[17] Rodriguez, K.L., Roemmich, R.T., Cam, B., Fregly, B.J. and Hass, C.J. (2013) Persons with Parkinson’s Disease Exhibit Decreased Neuromuscular Complexity during Gait. Clinical Neurophysiology, 124, 1390-1397. [Google Scholar] [CrossRef] [PubMed]
[18] 李亮亮, 陈香. 基于运动学和肌电特征的脑瘫爬行运动功能异常分析与评估[J]. 北京生物医学工程, 2018, 37(6): 566-574.
[19] Neumann, D.A., 诺伊曼. 骨骼肌肉功能解剖学[M]. 北京: 人民军医出版社, 2014.
[20] Boulay, C., Jacquemier, M., Pomero, V., Castanier, E., Authier, G., Chabrol, B., et al. (2014) Dynamic EMG of Peroneus Longus in Hemiplegic Children with Equinovarus. Annals of Physical and Rehabilitation Medicine, 57, 185-192. [Google Scholar] [CrossRef] [PubMed]
[21] Ippersiel, P., Dussault-Picard, C., Mohammadyari, S.G., De Carvalho, G.B., Chandran, V.D., Pal, S., et al. (2024) Muscle Coactivation during Gait in Children with and without Cerebral Palsy. Gait & Posture, 108, 110-116. [Google Scholar] [CrossRef] [PubMed]
[22] Falisse, A., Bar-On, L., Desloovere, K., Jonkers, I. and De Groote, F. (2018) A Spasticity Model Based on Feedback from Muscle Force Explains Muscle Activity during Passive Stretches and Gait in Children with Cerebral Palsy. PLOS ONE, 13, e0208811. [Google Scholar] [CrossRef] [PubMed]
[23] 张羽. 目标检测中的交并比损失函数研究[D]: [硕士学位论文]. 合肥: 安徽大学, 2022.