The clinical efficacy of acupuncture in the treatment of rheumatoid arthritis was evaluated by exploring the clinical efficacy of acupuncture in the treatment of rheumatoid arthritis in Chinese medicine. The study proposes a rheumatoid image recognition algorithm based on migration learning and EfficientNet, and the depth features extracted by the five mutually parallel convolutional layers in the model are used to obtain the respective AUC values by Monteiro cross-test to judge the ability of the convolutional layers to extract the depth features. Using the group test method, 80 test subjects were randomly divided into the treatment group and the control group, and the relevant clinical symptoms and routine test indexes (RF, CRP, ESR) before and after the treatment of the subjects in the two groups were detected, and the data of the two groups were statistically analyzed using SPSS24.0 statistical software. The results showed that the results obtained by rheumatoid image recognition technology on the test set were: SEN, SPE, LR+, and LR- were 82.6%, 77.98%, 3.75, and 0.22, respectively, with an AUC of 0.83. The feasibility of using image recognition algorithms for rheumatoid arthritis to predict and recognize joint images of patients has been verified. By observing the physical changes of the two groups before and after the test, it can be seen that after the treatment of the two groups of RA subjects, the improvement of various clinical symptom indexes in the treatment group was better than that in the control group (P0.05).