Application of Convolutional Neural Network Algorithm in Diagnosis of Chronic Cough and Tongue in Children with Traditional Chinese Medicine

Objective: This paper learns and studies the structure of convolutional neural network in deep learning, automatically extracts feature information, and explores the feasibility of this method in the classification model of chronic cough and tongue in children of traditional Chinese medicine, and assists in further objective analysis of tongue diagnosis of traditional Chinese medicine. Chemical. Through the research on the relationship between children's cough tongue and TCM syndrome type, severity of illness, disease course and laboratory examination, it provides objective basis for clinical syndrome differentiation of children with cough. Methods: The tongue images of children with cough and healthy children who met the inclusion criteria were collected and analyzed using DS01-B TCM tongue imager in the absence of interference factors such as diet and scraping. Images of tongue images on the 1st, 3rd, 5th, 7th, and 10th day from the observation date were collected, and the clinical symptoms, signs, TCM syndromes, mycoplasma antibodies, and blood routine results were recorded. The convolutional neural network algorithm is used to process the data level, including data deletion and tongue segmentation. Results: A total of 134 children with cough were collected as cough group and 30 healthy children were used as control group. The severity of the disease, the course of the disease, whether it is infected with cough and mycoplasma are closely related, and have nothing to do with blood routine. Conclusion: This study used DS01-B Chinese medicine tongue image instrument to collect and treat children's cough tongue image. The convolutional neural network algorithm was used to analyze the tongue image, which made the tongue image result more objective and provided an objective basis for clinical syndrome differentiation of children with cough.