During the diagnosis and treatment, the doctor will adopt different therapeutic regimens for different patient's conditions. As for the varied skill levels of doctors, it is a hard nut to crack in medicine and AI technology to assist them in obtaining updated and more effective therapeutic regimens in the course of diagnosis and treatment.

Traditional treatment recommendation and intervention technologies mainly use expert system solutions, but the problems of weak knowledge generalization ability and high update cost in this field need to be further solved. iFLYHealth purposely proposes a multi-level decision-making reasoning technology based on multi-source patient information and medical knowledge. Through the training of multi-dimensional information encoding of massive patient multi-source diagnosis and treatment information, medical knowledge graph of diagnosis and treatment, and clinical treatment cases, then through the multi-layer decision-making reasoning scheme, making full use of clinical information and medical knowledge, we finally achieve therapeutic regimen recommendation and early warning.