Symptom medicament association relation analysis method based on novel artificial intelligence technology

Abstract

The invention relates to a symptom medicament association relation analysis method based on novel artificial intelligence technology. The method comprises the steps of a symptom extracting module which extracts a symptom set S of a patient; a symptom word vector mapping module which maps a Chinese symptom set to a high-dimension dense word vector set; a symptom word vector coding module which utilizes a long-and-short-period memory network coding model for coding the symptom word vector set and generates an information vector C including total symptom information and n symptom coding information vector set HS; and a suggestion prescription generating module which utilizes the information vector C and the symptom coding information vector set HS as the input and utilizes a long-and-short-period memory network generating module which is combined with an attention model for generating a suggestion prescription which comprises L kinds of traditional Chinese medicines, namely an associationrelation between the symptoms and the traditional Chinese medicines. According to the method of the invention, from system input to system output, a clear logical structure and a clear hierarchy arerealized. Detailed system realization is obtained. An end-to-end (symptom end to traditional Chinese medicine end) system closed working mode is realized, thereby facilitating easy implementing and large-scale application.

Description

technical field [0001] The invention relates to two intersecting fields of artificial intelligence and traditional Chinese medicine, and specifically relates to an intelligent traditional Chinese medicine method that applies artificial intelligence to the field of traditional Chinese medicine to independently analyze the relationship between symptoms and medicines and generate results. Background technique [0002] At this stage, artificial intelligence is rising again and gradually entering all walks of life, and the field of traditional Chinese medicine has also entered the era of "artificial intelligence + Chinese medicine". Most of the existing intelligent TCM systems are based on the four diagnosis instruments of TCM and are based on expert systems and traditional machine learning methods, and realize intelligent TCM systems based on the experience of famous old Chinese medicine practitioners or ancient medical skills. This method mainly has the following deficiencies: ...

AI Technical Summary

Problems solved by technology

This method mainly has the following deficiencies: 1) The intelligent Chinese medicine realized by existing methods is based on rules and the rules are fixed and fixed by people, which requires a lot of human participation and cannot discover new laws and new experiences along with the process of practicing medicine

2) At present, the existing intelligent Chinese medicine does not use the ancient and modern actual medical records of traditional Chinese medicine as the knowledge of the intelligent Chinese medicine system, but these medical records generated in the actual practice of medicine have great medical value and analysis value of traditional Chinese medicine

3) Due to the implementation of existing intelligent Chinese medicine methods, the prescriptions generated by these methods are immutable, and there is no ability to discover new effective Chinese medicine combination ratios like Chinese medicine doctors

4) The existing intelligent Chinese medicine system cannot independently analyze and discover the relationship between symptoms and drugs, and this part of the work is still undertaken by humans; and some systems have complex structures, unclear logic, and are not suitable for large-scale applications

Images