Detection And Prediction Of Muscle Fatigue During Wheelchair Incremental Exercise Test

In recent years,the number of the wheelchair users are increasing.Muscle fatigue,particularly when it goes undetected,can put wheelchair users of assistive rehabilitation technologies at risk of serious injury,even irreversible damage.There has thus been considerable interest in the development of systems that can predict when and how muscles will fatigue.First,detect muscle fatigue by using ventilatory threshold and EMG fatigue threshold during the wheelchair incremental exercise test.Secondly,predict muscle fatigue by using support vector machine methods.Finally,quantify how muscle coordination changes over time(non-fatigue,transition-to-fatigue,and fatigue)during prolonged wheelchair propulsion.The main work and research results are as follows:(1)Detection of upper limb muscle fatigue of wheelchair users in single incremental exercise test.The ventilation threshold can be determined by the V-slope method,and the EMG thresholds were identified using the double-segment linear regression method.There is a good agreement between the EMG fatigue threshold and the ventilation threshold.The results of intra-class correlation coefficient(ICC)analysis showed that the EMG thresholds of the eight muscles were very similar(ICC = 0.91),indicating a single muscle EMG recording would be able to detect the onset of muscle fatigue.The EMG fatigue threshold method can serve as a valid and reliable tool for identifying the onset of muscular fatigue during wheelchair propulsion.(2)Prediction of upper limb muscle fatigue in wheelchair users during a single incremental exercise test.Support vector machines(SVM)was used to predict muscle fatigue,70% of the dataset were training sets and 30% of dataset were test sets.RBF kernel function was selected to establish a model for prediction.The accuracy,sensitivity and specificity were calculated by comparing the predicted results with the actual results.EMG signals from eight muscles generated the best predictors,with 97% accuracy.(3)Changes in coordination patterns of upper limb muscles during wheelchair incremental exercise test.Muscle coordination changes was analyzed by a combined use of wavelet analysis and principal component analysis that is capable of decomposing large EMG datasets into the summed activation of functional units of coordination in the form of synergies or coordinative structures.PC2 components show different levels and timing of muscle activation that vary with fatigue states.PC2 indicated a relative increase in BB and AD in the push phase(0-30% cycle)and more activity in the in recovery muscles,UT,MD,and PD,during the recovery phase at the transition to fatigue state.As for the fatigue state,PC2 are associated with increase in AD,PM,and IS activity in the late push phase to early recovery phase(30-60% cycle)and less activity in UT,MD,and PD during the recovery phase.PC3 components are associated with muscle co-contraction at fatigue state.A greater co-contraction of the shoulder joint AD and PD muscles during the early push phase and transitions(push to recovery and recovery to push),which may increase shoulder stiffness and accommodate the potential balance concerns.A rehabilitation program for wheelchair users should be based on a thorough understanding of shoulder muscle fatigue and coordination patterns change over time.By identifying EMG fatigue thresholds and coordination pattern changes,it is possible to predict when muscle fatigue will occur,which provides the foundation of an automated system.