计算机数学模型的远程医疗信息平台网络架构  

Network architecture of telemedicine information platform based on computer mathematical model

作　　者：李果[1] LI Guo(Children’s Hospital of Chongqing Medical University,Chongqing 400014,China)

机构地区：[1]重庆医科大学附属儿童医院,重庆400014

出　　处：《信息技术》2021年第8期133-138,144,共7页Information Technology

摘　　要：针对目前医疗信息存在大量孤岛、共享距离短等问题,该研究采用思维进化算法(MEA)、BP神经网络算法和区块链技术设计出一个远程医疗信息共享平台。由于传统的BP神经网络存在容易陷入局部最优和初始权值、阈值的随机性等缺陷,采用MEA算法来确定BP神经网络的初始权值和阈值,最终得到了医疗信息的融合模型。采用改进拜占庭协议的共识机制来实现网络节点共识,通过联盟区块链对医疗数据进行存储,保证了数据的安全性。试验结果表明,MEA+BP神经网络算法的精度相比传统BP神经网络提高了7.77%,稳定性提高了31.58%。Regrding the current medical information problems such as a large number of islands and short sharing distances,this research uses Mind Evolutionary Algorithm(MEA),BP neural network algorithms and Blockchain technology to design a telemedicine information sharing platform.Because the traditional BP neural network has the shortages of easy to fall into the local optimum and the randomness of initial weights and thresholds,the MEA is used to determine the initial weights and thresholds of the BP neural network,and finally a medical information fusion model is obtained.The consensus mechanism of the improved Byzantine agreement is adopted to realize the consensus of the network nodes,and the medical data is stored through the alliance Blockchain to ensure the security of the data.The test results show that the accuracy of the MEA+BP neural network algorithm increases by 7.77%and the stability increases by 31.58%compared with the traditional BP neural network.

关 键 词：计算机数学模型 远程医疗信息共享平台 MEA算法 BP神经网络 区块链 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]