MNCS.csv

In order to study the application of machine learning in myoelectric data, themachine learning method has been used for data mining and analysisso as tofind correlation characteristics. More than 2,300 myoelectric examination data from Sichuan Provincial Hospital of Traditional Chinese Medicine(TCM)for 10 monthshas been collected and recorded. By means of setting the inclusion criteria and excluding the irrelevant factors, the facial nerve electromyographyand auditory brainstem response test reports that meet the research criteriahave been screened out. Among them, there were 575 facial nerve electromyographyreports meeting the research criteria, and 233 auditory brainstem response (ABR) reportsin all. Based on these reports, the data sets have beenestablished.On the one hand, by comparing the advantages and disadvantages of several algorithms in EMG examination, the conclusions of random forest optimality are obtained. Meanwhile,the clinically obtained data areusedfor the interpretationandprediction. The results verify the clinical potential of machine learning in diagnosis and diagnostic assessment.On the other hand, the eigenvalues extracted by the random forest algorithm are adoptedto obtain the facial nerve electromyographyand the most important influencing factors of ABR, which facilitates the clinical analysis of doctors.