基于深度学习的智能中药处方识别与配伍禁忌分析系统OA

Objective To construct a system for recognizing traditional Chinese medicine prescriptions and analyzing incompatibility based on deep learning,thereby meeting the demands of digitalization and intelligence of traditional Chinese medicine prescriptions.Methods Firstly,a self-built dataset of traditional Chinese medicine prescriptions was constructed,and the image quality was improved through image binarization,denoising and super-resolution techniques.Secondly,the improved CTPN model and CRNN model were used to achieve precise and efficient detection and recognition of prescription text,and the association rule mining was combined to analyze the incompatibility of traditional Chinese medicine.Based on the theory of nature,taste,and channel tropism,the compatibility rules of traditional Chinese medicine were revealed to provide support for safe medication.Finally,an interactive system was built based on the Gradio framework to implement the functions of prescription image upload,text recognition,and compatibility analysis.Results The CTPN and CRNN models demonstrated good performance,achieving accuracy rates of 93.34%and 92.58%,respectively.Conclusion This traditional Chinese medicine prescription recognition and incompatibility analysis system based on deep learning performed well in the tasks of text detection and recognition and had the ability to assist in the discrimination of incompatibility.