To optimize the patient’s treatment process, to shorten the treatment time and moving path in the hospital, to help patients predict the time of each treatment process. Method/Process Based on graph theory and machine learning algorithm model, the intelligent treatment path recommendation model is constructed by integrating data information such as hospital spatial layout, diagnosis and treatment department distribution and patient treatment items. Based on the big data of hospital patients, the prediction model of patient treatment time is constructed to predict the time spent in registration, waiting for treatment or examination, etc. A real-time recommendation system for intelligent treatment paths is designed and implemented to provide patients with clear guidance and more accurate time-consuming estimation services. Result/Conclusion The intelligent real-time treatment path recommendation system integrates the capability of treatment time prediction and path recommendation, which significantly improves the treatment experience of patients and the service efficiency of hospitals.