Deep learning-based traditional Chinese medicine sublingual vein semantic segmentation extraction method and system

Abstract

The invention provides a deep learning-based traditional Chinese medicine sublingual vein semantic segmentation extraction method and system, and relates to the technical field of deep learning, and the method comprises the following steps: 1, carrying out the preprocessing of original data of a tongue body, and obtaining a data set; 2, inputting the data set into a semantic segmentation network for training and tongue segmentation prediction, and outputting a tongue segmentation prediction result; 3, inputting the tongue body segmentation result into a semantic segmentation network for training and sublingual vein segmentation prediction, and outputting a sublingual vein segmentation prediction result; and 4, performing post-processing on the tongue body segmentation prediction result and the sublingual vein segmentation prediction result. According to the invention, the doctor seeing cost of the traditional Chinese medicine can be reduced, the existing picture data is fully utilized, based on the current advanced machine learning technology, automatic segmentation of the tongue body and the sublingual collaterals is realized, the tongue body and the sublingual collaterals are extracted from the background, interference of other information is eliminated, and development of objective research of the tongue picture can be promoted.

Description

technical field [0001] The invention relates to the technical field of deep learning, in particular to a method and system for semantic segmentation and extraction of sublingual collaterals in traditional Chinese medicine based on deep learning. Background technique [0002] Tongue diagnosis is a non-invasive detection method that will not cause physical damage or radiation exposure to patients. However, it is greatly affected by various factors of subject and object, coupled with the lack of quantitative description, TCM diagnosis is usually relatively subjective, generally relying on the practical experience of clinicians, subjectivity is relatively high, and repeatability is poor, which has caused serious problems for TCM teaching and research. Many changes have been made, which has greatly affected the inheritance and development of tongue diagnosis. [0003] The objectification of tongue diagnosis is to use objective detection indicators, combined with tongue diagnosis...

AI Technical Summary

Problems solved by technology

However, it is greatly affected by various factors of subject and object, coupled with the lack of quantitative description, TCM diagnosis is usually relatively subjective, generally relying on the practical experience of clinicians, the subjectivity is relatively high, and the repeatability is poor, which has caused serious problems for TCM teaching and research. Many changes have been made, which has greatly affected the inheritance and development of tongue diagnosis.

Images