基于ASP.NET与深度学习的医学图像分割系统设计与实现OA

To address engineering problems in the integration of Deep Learning models and traditional Web information systems,this paper designs and implements a loosely coupled and highly available medical image segmentation information management system.The system adopts B/S architecture.It develops the user interface and business logic based on the ASP.NET framework at the front-end,and calls the independently deployed Mask R-CNN model service through RESTful API at the back-end to complete image segmentation tasks.It adopts SQL Server database for data management and designs an asynchronous task processing mechanism.It successfully builds a management system integrating user management,patient information management,image upload,and asynchronous segmentation task scheduling.Tests verify the effectiveness of the system architecture in function,performance,and heterogeneous service communication.This study provides a feasible scheme for solving engineering integration problems of Deep Learning models in medical information systems.The proposed loosely coupled architecture has good practicality and scalability.