柔性可拉伸材料引导下的多模态幼儿脑瘫识别方法研究

脑瘫（Cerebral Palsy, CP）是一种早期出现的运动和姿势发育障碍，当前脑瘫识别的主要方法为运动学量表结合病史，其存在着主观性强、识别效率低等问题，迫切需要一种客观、便捷的脑瘫识别方法。 CP幼儿的肌肉异常与运动障碍可以通过表面肌电信号（surface Electromyogram, sEMG）和惯性传感单元（Inertial Measurement Unit, IMU）表征，传统表面肌电电极普遍存在共形能力差、舒适性不足的问题，幼儿使用效果不尽如意。本文旨在制备一种共形能力高、舒适性强的柔性水凝胶电极，改善传统电极在幼儿群体使用时适配性不足的问题，同时探索一种基于多模态信号的CP识别方法，以期实现准确的 CP 识别。

研究中采用聚丙烯酰胺单体交联网络，制备了柔性水凝胶界面，结合可拉伸银浆与聚氨酯基底制备了柔性水凝胶电极，对其进行了拉伸、剥离、共形测试，研究表明所制备的电极具有良好的共形能力和舒适性，并通过了初步肌电信号采集验证。研究采集了6名脑瘫儿和10名健康儿的sEMG和IMU信号，分别构建了基于传统机器学习和基于MF-MobileNet（Multimodal Fusion MobileNet）的多模态融合脑瘫识别方法，并与常用深度学习模型进行对比验证，结果表明基于MF-MobileNet的多模态融合识别算法的准确率高于传统机器学习和其他深度学习模型，在留一验证中的脑瘫识别准确率达到了87.50%。初步验证了基于深度学习的sEMG和IMU多模态融合方法应用于CP识别的可行性，为临床CP识别提供了新的思路。

[1] ROSENBAUM P. A report: the definition and classification of cerebral palsy April 2006[J]. Developmental Medicine and Child Neurology, 2007, 49(6): 480-480.
[2] YANG S Y, XIA J Y, GAO J, et al. Increasing prevalence of cerebral palsy among children and adolescents in China 1988-2020: a systematic review and meta-analysis[J]. Journal of Rehabilitation Medicine, 2020, 53(5).
[3] HäGGLUND G, HOLLUNG S J, AHONEN M, et al. Treatment of spasticity in children and adolescents with cerebral palsy in Northern Europe: a CP-North registry study[J]. BMC Neurology, 2021, 21: 1-12.
[4] 高志平, 贾宁. 小儿脑瘫病因学研究进展[J]. 中国妇幼保健, 2012, 27(01): 149-150.
[5] 潘馨. 菏泽市小儿脑瘫临床分型调查及病因学研究[J]. 华北国防医药, 2009, 21(05): 12-14.
[6] 汪志国, 邱洪斌, 鲁向锋, 等. 小儿脑性瘫痪病因学的研究进展[J]. 疾病控制杂志, 2004(01): 52-55.
[7] FINCH-EDMONDSON M, MORGAN C, HUNT R W, et al. Emergent prophylactic, reparative and restorative brain interventions for infants born preterm with cerebral palsy[J]. Frontiers in Physiology, 2019, 10: 15.
[8] SPITTLE A J, MORGAN C, OLSEN J E, et al. Early diagnosis and treatment of cerebral palsy in children with a history of preterm birth[J]. Clinics in Perinatology, 2018, 45(3): 409-420.
[9] NOVAK I, MORGAN C, ADDE L. Early, accurate diagnosis and early intervention in cerebral palsy: advances in diagnosis and treatment[J]. JAMA Pediatrics, 2017, 171(9): 919-919.
[10] FJORTOFT T, EINSPIELER C, ADDE L, et al. Inter-observer reliability of the "Assessment of Motor Repertoire - 3 to 5 Months" based on video recordings of infants[J]. Early Human Development, 2009, 85(5): 297-302.
[11] MCCAY K D, HU P P, SHUM H P H, et al. A pose-based feature fusion and classification framework for the early prediction of cerebral palsy in infants[J]. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 2022, 30: 8-19.
[12] RIHAR A, MIHELJ M, PASIC J, et al. Infant trunk posture and arm movement assessment using pressure mattress, inertial and magnetic measurement units (IMUs)[J]. Journal of Neuroengineering and Rehabilitation, 2014, 11: 1-14.
[13] XIONG Q L, WU X Y, YAO J, et al. Inter-limb muscle synergies and kinematic analysis of hands-and-knees crawling in typically developing infants and infants with developmental delay[J]. Frontiers in Neurology, 2018, 9.
[14] DI N F, STRAZZA A, MENGARELLI A, et al. EMG-based characterization of walking asymmetry in children with mild hemiplegic cerebral palsy[J]. Biosensors-Basel, 2019, 9(3): 82.
[15] KUO C C, HUANG H P, WANG T M, et al. Tendon release reduced joint stiffness with unaltered leg stiffness during gait in spastic diplegic cerebral palsy[J]. Plos One, 2021, 16(1): 610-616.
[16] WAGENAAR N, VERHAGE C H, DE VRIES L S, et al. Early prediction of unilateral cerebral palsy in infants at risk: MRI versus the hand assessment for infants[J]. Pediatric Research, 2020, 87(5): 932-939.
[17] 陈世煌, 陈杰云, 黄日升, et al. 多模态功能磁共振成像技术诊断儿童脑瘫的应用价值[J]. 基层医学论坛, 2024, 28(01): 106-108.
[18] APOSTOLIDIS K D, PAPAKOSTAS G A. A survey on adversarial deep learning robustness in medical image analysis[J]. Electronics, 2021, 10(17): 21-32.
[19] PRECHTL H F R. State of the art of a new functional assessment of the young nervous system. An early predictor of cerebral palsy[J]. Early Human Development, 1997, 50(1): 1-11.
[20] EINSPIELER C, PRECHTL H F R, FERRARI F, et al. The qualitative assessment of general movements in preterm, term and young infants - review of the methodology[J]. Early Human Development, 1997, 50(1): 47-60.
[21] 杨红, 邵肖梅. 全身运动质量评估[J]. 中国循证儿科杂志, 2007(02): 138-143.
[22] HADDERSALGRA M. The assessment of general movements is a valuable technique for the detection of brain dysfunction in young infants. A review[J]. Acta Paediatrica, 1996, 85: 39-43.
[23] GUIMARAES C L N, REINAUX C M, BOTELHO A C G, et al. Motor development evaluated by Test of Infant Motor Performance: comparison between preterm and full-term infants[J]. Brazilian Journal of Physical Therapy, 2011, 15(5): 357-362.
[24] 刘鹏, 黄东锋, 江沁, et al. 脑瘫患儿粗大运动功能测量量表的标准化研究[J]. 中国康复医学杂志, 2004(03): 10-13.
[25] CARNAHAN K D, ARNER M, HäGGLUND G. Association between gross motor function (GMFCS) and manual ability (MACS) in children with cerebral palsy. A population-based study of 359 children[J]. BMC Musculoskeletal Disorders, 2007, 8: 1-7.
[26] WU Y C, VAN R I M, BUURMAN M T, et al. Temporal and spatial localisation of general movement complexity and variation-Why Gestalt assessment requires experience[J]. Acta Paediatrica, 2021, 110(1): 290-300.
[27] MARCROFT C, KHAN A, EMBLETON N D, et al. Movement recognition technology as a method of assessing spontaneous general movements in high risk infants[J]. Frontiers in Neurology, 2015, 5: 111-125.
[28] BACCINELLI W, BULGHERONI M, SIMONETTI V, et al. Movidea: a software package for automatic video analysis of movements in infants at risk for neurodevelopmental disorders[J]. Brain Sciences, 2020, 10(4): 203.
[29] 李洪华, 单玲, 王冰, et al. 运动识别技术在评估早产儿自发性全身运动中的应用[J]. 中国当代儿科杂志, 2017, 19(12): 1306-1311.
[30] MüNZNER S, SCHMIDT P, REISS A, et al. CNN-based sensor fusion techniques for multimodal human activity recognition[J]. Proceedings of the 2017 ACM International Symposium on Wearable Computers, 2017: 158-165.
[31] COONEY N J, MINHAS A S. Humanoid robot based platform to evaluate the efficacy of using inertial sensors for spasticity assessment in cerebral palsy[J]. IEEE Journal of Biomedical and Health Informatics, 2022, 26(1): 254-263.
[32] GROOS D, ADDE L, AUBERT S, et al. Development and validation of a deep learning method to predict cerebral palsy from spontaneous movements in infants at high risk[J]. JAMA Network Open, 2022, 5(7): 1325.
[33] LEONARD C T, HIRSCHFELD H, FORSSBERG H. The development of independent walking in children with cerebral-palsy[J]. Developmental Medicine and Child Neurology, 1991, 33(7): 567-577.
[34] SARCHER A, BROCHARD S, PERROUIN-VERBE B, et al. Detection of pronator muscle overactivity in children with unilateral spastic cerebral palsy: Development of a semi-automatic method using EMG data[J]. Annals of Physical and Rehabilitation Medicine, 2019, 62(6): 409-417.
[35] 熊启亮. 婴幼儿膝爬运动功能发育状态对肌肉收缩及关节运动协同模式的影响[D]. 重庆大学, 2019.
[36] 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(06): 566-574.
[37] FUGLSANG-FREDERIKSEN A. The role of different EMG methods in evaluating myopathy[J]. Clin Neurophysiol, 2006, 117(6): 1173-1189.
[38] 李梅. 表面肌电信号的信息熵分析[D]. 西安电子科技大学, 2023.
[39] SUN Y, REN L, JIANG L, et al. Fabrication of composite microneedle array electrode for temperature and bio-signal monitoring[J]. Sensors, 2018, 18(4): 1193.
[40] KRACHUNOV S, CASSON A J. 3D printed dry EEG electrodes[J]. Sensors, 2016, 16(10): 1635.
[41] LOZANO J, STOEBER B. Fabrication and characterization of a microneedle array electrode with flexible backing for biosignal monitoring[J]. Biomedical Microdevices, 2021, 23: 1-19.
[42] ALIZADEH-MEGHRAZI M, YING B, SCHLUMS A, et al. Evaluation of dry textile electrodes for long-term electrocardiographic monitoring[J]. Biomedical Engineering Online, 2021, 20: 1-20.
[43] TYBRANDT K, KHODAGHOLY D, DIELACHER B, et al. High‐density stretchable electrode grids for chronic neural recording[J]. Advanced Materials, 2018, 30(15): 1706520.
[44] ARQUILLA K, WEBB A K, ANDERSON A P. Textile electrocardiogram (ECG) electrodes for wearable health monitoring[J]. Sensors, 2020, 20(4): 1013.
[45] LA T G, QIU S, SCOTT D K, et al. Two‐layered and stretchable e‐textile patches for wearable healthcare electronics[J]. Advanced healthcare materials, 2018, 7(22): 1801033.
[46] SHAY T, VELEV O D, DICKEY M D. Soft electrodes combining hydrogel and liquid metal[J]. Soft Matter, 2018, 14(17): 3296-3303.
[47] NAGAMINE K, CHIHARA S, KAI H, et al. Totally shape-conformable electrode/hydrogel composite for on-skin electrophysiological measurements[J]. Sensors and Actuators B: Chemical, 2016, 237: 49-53.
[48] YUAN M, LONG Y, LIU T, et al. Soft electronics for advanced infant monitoring[J]. Materials Today, 2024.
[49] LIU H, CHEN X, WANG Z, et al. Development of soft dry electrodes: from materials to structure design[J]. Interfaces, 2023, 10: 11.
[50] YE S F, ZHU K H, MA W B, et al. Conducting polymer hydrogel driven by sodium chloride as high performance flexible supercapacitor electrode[J]. Journal of the Electrochemical Society, 2022, 169(7): 73501.
[51] SHEN G C, GAO K P, ZHAO N, et al. A fully flexible hydrogel electrode for daily EEG monitoring[J]. IEEE Sensors Journal, 2022, 22(13): 12522-12529.
[52] 孙静, 李韩飞, 郭培志, et al. 柔性可拉伸导电材料用于生理信号获取与反馈的研究简述[J]. 材料导报, 2021, 35(05): 5158-5165.
[53] STEGEMAN D, HERMENS H. Standards for surface electromyography: The European project Surface EMG for non-invasive assessment of muscles (SENIAM)[J]. Enschede: Roessingh Research and Development, 2007, 10: 8-12.
[54] CHOWDHURY R H, REAZ M B I, ALI M A B M, et al. Surface electromyography signal processing and classification techniques[J]. Sensors, 2013, 13(9): 12431-12466.
[55] BURHAN N, GHAZALI R. Feature extraction of surface electromyography (sEMG) and signal processing technique in wavelet transform: A review[C]//2016 IEEE International Conference on Automatic Control and Intelligent Systems (I2CACIS). IEEE, 2016: 141-146.
[56] ILLAVARASON P, RENJIT J A, KUMAR P M. Medical diagnosis of cerebral palsy rehabilitation using eye images in machine learning techniques[J]. Journal of Medical Systems, 2019, 43(8): 278.
[57] IHLEN E A F, STOEN R, BOSWELL L, et al. Machine learning of infant spontaneous movements for the early prediction of cerebral palsy: a multi-site cohort study[J]. Journal of Clinical Medicine, 2020, 9(1): 5.
[58] GOODLICH B I, ARMSTRONG E L, HORAN S A, et al. Machine learning to quantify habitual physical activity in children with cerebral palsy[J]. Developmental Medicine and Child Neurology, 2020, 62(9): 1054-1060.
[59] KANIYAMATTAM K, VEETTIL V, RIVERA M E, et al. Design and development mechanics of a random forest-based decision-tree model for sub-acute rumen acidosis management in feedlots[J]. Journal of Animal Science, 2023, 101 (Supplement_3): 55-56.
[60] BREIMAN L. Random forests[J]. Machine Learning, 2001, 45(1): 5-32.
[61] CUNNINGHAM P, DELANY S J. K-nearest neighbour classifiers - a tutorial[J]. ACM Computing Surveys, 2021, 54(6): 1-25.
[62] FAN R-E, CHANG K-W, HSIEH C-J, et al. LIBLINEAR: A Library for Large Linear Classification[J]. J Mach Learn Res, 2008, 9: 1871–1874.
[63] WANG L, HAN M, LI X J, et al. Review of classification methods on unbalanced data sets[J]. IEEE Access, 2021, 9: 64606-64628.
[64] ÖZATES M E, KARABULUT D, SALAMI F, et al. Predicting joint moments from kinematics in patients with cerebral palsy using deep learning[J]. Gait & Posture, 2023, 100: 28-29.
[65] KOLAGHASSI R, MARCELLI G, SIRLANTZIS K. Deep learning models for stable gait prediction applied to exoskeleton reference trajectories for children with cerebral palsy[J]. IEEE Access, 2023, 11: 31962-31976.
[66] KIM Y K, VISSCHER R M S, VIEHWEGER E, et al. A deep-learning approach for automatically detecting gait-events based on foot-marker kinematics in children with cerebral palsy-Which markers work best for which gait patterns?[J]. Plos One, 2022, 17(10): 275878.
[67] LI Z W, LIU F, YANG W J, et al. A survey of convolutional neural networks: analysis, applications, and prospects[J]. IEEE Transactions on Neural Networks and Learning Systems, 2022, 33(12): 6999-7019.
[68] ZHOU Y, CHEN S C, WANG Y M, et al. Review of research on lightweight convolutional neural networks[J]. Proceedings of 2020 IEEE 5th Information Technology and Mechatronics Engineering Conference (ITOEC 2020), 2020: 1713-1720.
[69] XU L K, ZHANG K Q, YANG G K, et al. Gesture recognition using dual-stream CNN based on fusion of sEMG energy kernel phase portrait and IMU amplitude image[J]. Biomedical Signal Processing and Control, 2022, 73: 103364.
[70] GUO W C, SHENG X J, LIU J W, et al. Towards zero training for myoelectric control based on a wearable wireless sEMG armband[J]. 2015 IEEE/Asme International Conference on Advanced Intelligent Mechatronics (AIM), 2015: 196-201.
[71] ZHEN T, ZHENG H, YAN L. Human motion mode recognition based on multi-parameter fusion of wearable inertial module unit and flexible pressure sensor[J]. Sensors and Materials, 2022, 34(3): 1017-1031.
[72] GROOS D, ADDE L, AUBERT S, et al. Development and validation of a deep learning method to predict cerebral palsy from spontaneous movements in infants at high risk[J]. JAMA Network Open, 2022, 5(7): 21325.
[73] MACHIREDDY A, VAN S J, WILSON J L, et al. A video/IMU hybrid system for movement estimation in infants[C]//2017 39th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC). IEEE, 2017: 730-733.