Font Size: a A A

ICA-based Study On Muscle-tendon Units Localization And Muscle Force Estimation

Posted on:2018-11-16Degree:MasterType:Thesis
Country:ChinaCandidate:S P WangFull Text:PDF
GTID:2347330512486705Subject:Biomedical engineering
Abstract/Summary:PDF Full Text Request
Sport biomechanics has an important bearing on people's movement health,as it can help understand the inherent mechanisms of movement.Human body movement can be described by a neuromusculoskeletal model,within which the skeletal muscles are considered as basic units to provide momentum.Therefore,the study of sport biomechanics should be based on a clear understanding of the physiological characteristics of skeletal muscles.A single skeletal muscle can be further divided into muscle tendon units(MTUs).For different skeletal muscles,the numbers of MTUs vary and the activation of one MTU differs from that of another in the process of movement.Surface electromyography signals(sEMG),which are collected from the surface skin in the process of skeletal muscle contractions,are able to record information about muscle contractions.Furthermore,the sEMG recording is recommended due to its noninvasive and convenient characters.Compared with conventional surface electrodes,the advanced high-density electrode array can help discriminate superposed waveforms of the motor unit action potentials in the recorded sEMG,thus improving the accuracy of the quantitative sEMG analysis.On this basis,the high-density sEMG recording provides an effective way of locating MTUs of skeletal muscles,analyzing activation level and estimating muscle force.In the thesis,the high-density sEMG signals processed by independent component analysis(ICA)were used to locate MTUs,analyze activation levels of the MTUs and estimate the muscle force.The main research contents and results can be summarized as follows:(1)Location of MTUs and their activation level analysis.A method based on ICA algorithm was proposed to locate MTUs and analyze their activation level.Firstly,the high-density electrode array was selected to collect sEMG signals of gastrocnemius muscle and extensor digitorum communis muscle.Secondly,through ICA algorithm,the high-density sEMG signals were decomposed into a weight matrix and a source matrix.On this basis,the resultant source matrix obtained by the ICA contains sEMG sources(representing MTU activities)and noise sources and their discrimination was realized based on spectrum analysis.Then,the source signals of MTUs were identified and reserved.Finally,the location of each MTU was determined by the spatial distribution of the corresponding sEMG source in the weight matrix,and its activation level could accordingly obtained through the resultant sources.Eight subjects were recruited in the experiment and sEMG data were obtained from GM and EDC,during the subject's performance of calf raise and individual finger extension,respectively.The experimental results showed that the method proposed in the thesis is able to locate individual MTUs and analyze their activation levels effectively during dynamic tasks.Meanwhile,the method based on ICA was superior to the routine method based on EMG amplitude.(2)Muscle force estimation based on MTUs.Source signals of MTUs separated by ICA were processed and used as input signals of EMG-muscle force estimation model(it was actually a neural network model in this thesis),which promoted the study of muscle force estimation to the MTU level.Biceps brachii was selected as the target muscle.The two modes of exerting muscle force,namely periodic constant force and linear lift force,and two postures,namely neutral and supination postures,were designed and involved in the elbow flexion task.Eight subjects were required to perform tasks at the 20%,40%and 60%maxium voluntary contraction(MVC)force levels.The experimental results showed that the muscle force estimation based on MTUs was able to select appropriate source signals as input to the muscle force estimation model across different tasks.Compared with the method based on raw EMG signals,the proposed MTU-based method was capable of obtaining higher accuracy in muscle force estimation.Sport biomechanics is closely related to human life,and therefore it is essential to make efforts in this field.The study of MTUs provides better understandings of the sport biomechanics and it also promotes the development of sports biomechanics.Furthermore,it also offers a new way to examine the functional states of MTUs through dynamic tasks and to accurately estimate the muscle force,and conducts sport biomechanics research from the muscle level to the MTU level.
Keywords/Search Tags:Sport biomechanics, Muscle tendon unit, Surface electromyography, High density electrode array, Independent component analysis, Muscle force estimation
PDF Full Text Request
Related items