Hospital guidance method and system

Abstract

The embodiment of the invention provides a hospital guide method. The method comprises the following steps of: gradually reasoning symptoms input by a patient based on a medical knowledge graph, and determining a first result of a hospital guide department and a potential disease; inputting the symptoms input by the patient into an end-to-end classification model based on deep learning, and outputting a second result of the hospital guide department and the potential disease; based on the first result and the second result, outputting an effective hospital guide department and potential disease to the patient; providing an interpretable inference path corresponding to the effective hospital guide department and the potential disease based on the medical knowledge graph. The embodiment of the invention further provides a hospital guide system. According to the method and system of the embodiments of the invention, a knowledge reasoning function is added in hospital guidance, and end-to-end classification and a reasoning model are combined, so that high recall and high precision of an analytic expression of end-to-end classification are fully exerted, the accuracy of hospital guidance is improved; and meanwhile, the interpretable path and related information recommendation of hospital guidance are given, and the use experience of the patient is improved.

Description

technical field [0001] The invention relates to the field of intelligent voice, in particular to a diagnosis guidance method and system. Background technique [0002] With the development of intelligent voice interaction, feedback is given by recognizing the user's intention from the user's input. In the hospital, patients may not know which department to go to. At this time, there are professional staff to help patients guide the diagnosis. However, with the advancement of technology, online registration / seeing a doctor / prescribing medicine is gradually popularized. At this time, there is usually no professional staff. The staff helped, and some patients did not know what number to register for their symptoms. In order to facilitate the accurate registration of patients, the existing technology usually uses: [0003] The end-to-end classification method of deep learning: obtain a large amount of training data by labeling the department label of the chief complaint, use th...

AI Technical Summary

Problems solved by technology

[0008] In order to at least solve the problem that the existing guidance methods in the prior art lack common sense knowledge and are prone to obvious wrong results. In the voice interaction, the patient's oral expression is not clear, which will also affect the expression matching effect.

Images