Risk prediction of hypertension complications based on the intelligent algorithm optimized Bayesian network

Hypertension and its related complications could be a major threat issue for cardiopathy and stroke. Effective prevention and control can decrease the incidence rate of complications in hypertension. Based on the medical data of 3062 patients with cardiovascular and cerebrovascular diseases from 201...

Full description

Saved in:
Bibliographic Details
Published in:Journal of combinatorial optimization Vol. 42; no. 4; pp. 966 - 987
Main Authors: Du, Gang, Liang, Xi, Ouyang, Xiaoling, Wang, Chunming
Format: Journal Article
Language:English
Published: New York Springer US 01.11.2021
Springer Nature B.V
Subjects:
Algorithms
Back propagation networks
Bayesian analysis
Combinatorics
Convex and Discrete Geometry
Data mining
Decision trees
Hypertension
Mathematical Modeling and Industrial Mathematics
Mathematics
Mathematics and Statistics
Neural networks
Operations Research/Decision Theory
Optimization
Particle swarm optimization
Risk analysis
Theory of Computation
Improved particle swarm optimization
Risk prediction
Hypertension complications
Intelligent algorithm optimized Bayesian network
ISSN:1382-6905, 1573-2886
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Hypertension and its related complications could be a major threat issue for cardiopathy and stroke. Effective prevention and control can decrease the incidence rate of complications in hypertension. Based on the medical data of 3062 patients with cardiovascular and cerebrovascular diseases from 2017 to 2018 in a grade-A tertiary hospital in Shanghai, the study identified the risk factors of hypertension complications by text mining. On this basis, the K2 algorithm based on the improved particle swarm optimization was proposed to optimize the structure of the Bayesian network (BN) by establishing a multi-population cooperative search mechanism. Then the optimized BN was used to analyze and predict the incidence rate of hypertension complications. Results indicate that the major indicators of accuracy, sensitivity, specificity, and AUC have been improved, and the proposed algorithm is superior to the common data mining algorithms such as the back propagation neural network and the decision tree. Through the proposed model and algorithm, the high-risk factors were identified and the occurrence probability of hypertension complications was predicted, which could provide the personalized health management guidance for hypertensive patients to prevent and control hypertension complications.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:1382-6905
1573-2886
DOI:10.1007/s10878-019-00485-z