Method for assisting traditional Chinese medicine intelligent diagnosis based on face video

Abstract

The invention discloses an auxiliary traditional Chinese medicine intelligent diagnosis method based on a face video, and the method comprises the following steps: (1) recording the face video in realtime through employing a smartphone camera, and outputting a face front picture, a tongue upper part picture, a tongue lower part picture and a face front one-minute video; (2) recognizing facial symptoms according to the front picture of the human face, and recognizing tongue symptoms according to the tongue picture; recognizing pulse symptoms according to the face video; (3) according to the facial symptom, tongue symptom and pulse symptom information recognized by the face video image, analyzing the association strength of the known symptom set in the medical knowledge graph, reasoning other symptoms possibly existing in the user, and obtaining a final symptom set S; and (4) calculating the association strength between the symptom set S and all syndromes according to the collected symptom set S, and inferring the closest physique tendency type and suspected disease name according to the corresponding relationship between the syndromes and physiques and diseases.

Description

technical field [0001] The invention relates to the technical field of TCM disease diagnosis, in particular to a method for assisting TCM intelligent diagnosis based on human face video. Background technique [0002] Artificial intelligence-assisted diagnosis and treatment reasoning refers to allowing computers to learn medical knowledge, simulate doctors' thinking and diagnostic reasoning methods, and give reliable diagnosis and treatment plans. In diagnosis, artificial intelligence needs to obtain the patient's disease information, reason and judge the cause and development trend of the disease through the learned medical knowledge, and form a treatment plan. [0003] Artificial intelligence-assisted reasoning technology can be applied to all stages of diagnosis and treatment. Among them, self-service diagnosis and treatment of sub-health is one of the application directions. Sub-health self-service diagnosis and treatment refers to the use of artificial intelligence techn...

AI Technical Summary

Problems solved by technology

Without the participation of actual doctors, the current AI-assisted diagnosis and treatment system collects symptoms in the following ways: the first is that the user chooses to input symptoms by himself, and then the system makes a reasoning diagnosis. This method is limited by the user's medical knowledge. The level of knowledge is highly subjective, and the results of reasoning and diagnosis are also highly subjective and less accurate.

The second is the question-and-answer interaction method. The user inputs any symptom, and the system infers the most suitable symptom question and answer to achieve more complete symptom data collection. This method requires multiple question-and-answer interactions, and the symptom collection takes a long time. low experience

The third is face-to-face and tongue-diagnosis symptom input based on image recognition. This method is limited by the fact that the face-to-face and tongue-diagnosis parameters acquired by image recognition are few. If combined with consultation, there is also subjectivity and long question-and-answer interaction time. The problem

Images