Multi-center biomedical data co-processing system and method without patient data sharing

Abstract

The invention discloses a multi-center biomedical data co-processing system and method without patient data sharing. The multi-center biomedical data co-processing system comprises a cloud server anda medical center client, wherein the cloud server is used for coordinating various medical center model parameters and asynchronously calculating; the medical center client is used for carrying out local high-performance calculation on data; the cloud server comprises a parameter manager and a task scheduler. A parameter sharing mechanism and an asynchronous communication mechanism are connected and managed through a parameter manager and a task scheduler of a cloud server and are jointly applied to multi-center medical data collaborative computing; security isolation between medical center data and a cloud server can be met, and the privacy of medical center patient data is fully protected; meanwhile, compared with a direct synchronous exchange data calculation result, the calculation waiting time is effectively shortened, and the analysis efficiency and the data processing capacity of multi-center cooperative processing are greatly improved.

Description

technical field [0001] The invention belongs to the field of data collaborative processing, and in particular relates to a multi-center biomedical data collaborative processing system and method without patient data sharing. Background technique [0002] The development of big data technology and the analysis and research of big data play a supporting role in all aspects of social development and progress. With the development of modern digital technology, the informatization of hospital diagnosis and treatment process and the implementation of electronic medical records, massive data will be generated in the medical process. The biomedical data generated globally every year reaches exabyte level, but the data of each medical center is independent of each other, and the patient population has great heterogeneity. Therefore, the Department of Biomedical Research encourages the collaborative processing of medical big data with overall relevance. [0003] In the existing multi...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to propose a new multi-center medical data collaborative processing system and method without patient data sharing in order to solve the problems of user data privacy leakage and synchronous update model waiting time in the existing multi-center medical data collaboration process. Facing the actual application scenario, the present invention uses a distributed collaborative processing method to solve the problem of medical big data analysis on the premise of fully avoiding the risk of external circulation and leakage of all patient data, and completes the collaborative optimization and optimization of the model through the method of asynchronous sharing of the model. Improvement of computing efficiency

Images