Research on Development and Integration of Artificial Intelligence Aided Diagnosis System

Abstract

In view of the current lack of agile development project management and steep learning curve of tensorflow serving deployment in the medical field of artificial intelligence, based on the named entity recognition and entity annotation of Chinese electronic medical records, this paper uses the iterated dilated convolutional neural networks(ID-CNNs) and conditional random field modeling of deep learning, develops restful API by combining the model training results with Python flask, and uses spring boot / spring cloud Alibaba, Using container technology docker and enterprise kubernetes management platform rancher, using agile development project management and the idea of front-end and back-end separation, promote the establishment of deep learning model, the parallel development of training process and system software, and the agile development, integration, debugging and deployment of artificial intelligence aided diagnosis system.