疾病自动诊断是智慧医疗的发展趋势。为实现高质量的临床应用，核心问题是探究造成数据源多样性的源端因子对诊断质量的影响，确定高精度的长短期疾病诊断模型评价标准。本项目将充分利用心脏疾病特征，通过消除源端因子对模型质量的影响，优化目标调整模型结构，构建基于三个完整心拍构建的实时自动分类模型；利用分类模型对长期监护信息准确标注以建立长短期预测机制，量化患病风险构建疾病动态预测体系，生成具有特异性的心电数据库；结合迁移学习理论泛化自动分类模型，在若干医疗图像领域开展模型适用性实证研究，为模型的迁移提供实践支撑。本研究集中探讨建立各类源端因子对模型的影响体系，关注长期诊断与预测模型质量的动态交互关系，量化医疗数据差异提高迁移参数适应性。较之于传统方法，本项目构建的模型将为心脏疾病的自动诊断与长期监护带来新思路，也同时为其他病症的智能医疗诊断策略提供理论指导意义。

Automatic classification and diagnosis of diseases is a significant trend of intelligent medical. Aiming at delivering high quality clinical application, a key issue is to explore the influence of source factors that cause the diversity of data sources on the quality of diagnosis, and to determine the evaluation criteria of high-precision long-term and short-term disease diagnosis models.. This project will make full use of the features of heart disease to build a real-time automatic classification model based on three beats and adjust the model structure by eliminating the influence of source factors as an optimization target. It will accurately mark long-term monitoring information based on the classification model, establish a long-term and short-term prediction mechanism, quantify the risk of disease, build a dynamic disease prediction system, and generate a patient-specific database called FZU-FPH. The project will combine the migration learning theory to generalize the automatic classification model and conduct case verification in other medical image fields to provide practical support for the migration of model framework. Compared with the traditional model, this study focused on establishing the influence system of various source factors on the model, paying close attention to the dynamic interaction between the quality of long-term diagnosis and prediction model, and quantifying the degree of difference of medical data to improve the adaptability of migration parameters. The model constructed by this project can bring new ideas to the intelligent diagnosis and treatment of heart diseases. At the same time, it has a broad theoretical guiding significance to the intelligent medical strategies for other diseases, and further alleviate the social contradictions caused by the lack of medical resources.