Utilizing machine learning for nerve conduction velocity analysis

Abstract Nerve conduction (NC) studies often utilized electrodiagnostic tests for assessing the conduction of both motor and sensory nerves. In a cohort study of 216 individuals within the age range of 20 to 59 years, from the findings of the Poursina Electrodiagnostic Center, Rasht, Iran, healthy subjects underwent machine learning (ML) techniques to evaluate the conduction of nerves situated in the upper extremities, specifically the median, ulnar, and radial nerves. Through this investigation, normal values were successfully set up. There are multiple variables, such as gender, age, weight, and height that can influence NC. Given the historical background and thorough medical assessment, assuming the absence of any NC agents or underlying medical conditions. The results showed the successful establishment of normal values for NC and the favorable performance of the ML models in predicting NC values.