In the field of medical education and training, there are still significant challenges in the decomposition and mastery of complex and fine hand manipulation skills. The work aims to propose a method to encode the fine movements of doctors' hands by combining the subjective judgment information with surface electromyographic signals. Firstly, the definition of behavior coding and the characteristics of hand fine movements were clarified, and the application path of EMG in behavior coding was sorted out. Based on the above research background and foundation, the acquisition of subjective judgment information and the construction of coding model were completed, and the user experiment to measure sEMG signal was designed and implemented. The fine movement coding of doctors' hands was verified through node judgment and movement segmentation of EMG data. The behavior coding related to fine hand movement skills provides a structured behavior sequence method combining surface electromyography and subjective judgment, which not only helps medical students understand and learn standardized surgical behaviors, but also provides an empirical reference for the development, training and testing of surgical robots in the future.