Prediction of High-Risk Cardiac Arrhythmia Based on Optimized Deep Active Learning

In contemporary society, numerous challenges are effectively addressed by applying computer science and artificial intelligence. Each year, a substantial number of fatalities occur due to various illnesses; however, many of these can be anticipated and managed through the utilization of AI technolog...

Full description

Saved in:
Bibliographic Details
Main Authors: Homeyra Amiri, Javad Mohammadzadeh, Seyed Mohsen Mirhosseini, Alireza Nikravanshalmani
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10900369/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850028722007048192
author Homeyra Amiri
Javad Mohammadzadeh
Seyed Mohsen Mirhosseini
Alireza Nikravanshalmani
author_facet Homeyra Amiri
Javad Mohammadzadeh
Seyed Mohsen Mirhosseini
Alireza Nikravanshalmani
author_sort Homeyra Amiri
collection DOAJ
description In contemporary society, numerous challenges are effectively addressed by applying computer science and artificial intelligence. Each year, a substantial number of fatalities occur due to various illnesses; however, many of these can be anticipated and managed through the utilization of AI technologies. Among the significant health concerns, cardiac arrhythmia is particularly noteworthy within the domain of cardiovascular diseases, which is the primary focus of this study. Cardiac arrhythmias are classified into distinct categories, with some being benign while others present considerable risks. This article proposes a novel methodology that employs advanced AI techniques for predicting high-risk cases of cardiac arrhythmia in individuals. The study delineates a new framework that integrates deep learning, fuzzy logic, and active learning, aiming to provide an innovative approach for the early detection of cardiac arrhythmias through Optimized Deep Active Learning (ODAL). The proposed ODAL framework achieved superior performance, with an F1 score of 90% for non-sinus rhythm classification and a precision of 92%. Additionally, the ROC analysis of the framework demonstrated satisfactory performance, achieving an accuracy of 86%, sensitivity of 86%, and specificity of 82%. These results underscore the model’s effectiveness in accurately identifying high-risk arrhythmia cases compared to other classifiers.
format Article
id doaj-art-58ae593bcce843a5960aaeaa111abb3e
institution DOAJ
issn 2169-3536
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-58ae593bcce843a5960aaeaa111abb3e2025-08-20T02:59:45ZengIEEEIEEE Access2169-35362025-01-0113390063903210.1109/ACCESS.2025.354490410900369Prediction of High-Risk Cardiac Arrhythmia Based on Optimized Deep Active LearningHomeyra Amiri0Javad Mohammadzadeh1https://orcid.org/0000-0003-1889-0294Seyed Mohsen Mirhosseini2https://orcid.org/0000-0002-2990-9598Alireza Nikravanshalmani3https://orcid.org/0000-0003-1308-7072Department of Computer Engineering, Karaj Branch, Islamic Azad University, Karaj, IranDepartment of Computer Engineering, Karaj Branch, Islamic Azad University, Karaj, IranDepartment of Computer Engineering, Karaj Branch, Islamic Azad University, Karaj, IranDepartment of Computer Engineering, Karaj Branch, Islamic Azad University, Karaj, IranIn contemporary society, numerous challenges are effectively addressed by applying computer science and artificial intelligence. Each year, a substantial number of fatalities occur due to various illnesses; however, many of these can be anticipated and managed through the utilization of AI technologies. Among the significant health concerns, cardiac arrhythmia is particularly noteworthy within the domain of cardiovascular diseases, which is the primary focus of this study. Cardiac arrhythmias are classified into distinct categories, with some being benign while others present considerable risks. This article proposes a novel methodology that employs advanced AI techniques for predicting high-risk cases of cardiac arrhythmia in individuals. The study delineates a new framework that integrates deep learning, fuzzy logic, and active learning, aiming to provide an innovative approach for the early detection of cardiac arrhythmias through Optimized Deep Active Learning (ODAL). The proposed ODAL framework achieved superior performance, with an F1 score of 90% for non-sinus rhythm classification and a precision of 92%. Additionally, the ROC analysis of the framework demonstrated satisfactory performance, achieving an accuracy of 86%, sensitivity of 86%, and specificity of 82%. These results underscore the model’s effectiveness in accurately identifying high-risk arrhythmia cases compared to other classifiers.https://ieeexplore.ieee.org/document/10900369/Deep learningactive learningcardiac arrhythmiafuzzy logic
spellingShingle Homeyra Amiri
Javad Mohammadzadeh
Seyed Mohsen Mirhosseini
Alireza Nikravanshalmani
Prediction of High-Risk Cardiac Arrhythmia Based on Optimized Deep Active Learning
IEEE Access
Deep learning
active learning
cardiac arrhythmia
fuzzy logic
title Prediction of High-Risk Cardiac Arrhythmia Based on Optimized Deep Active Learning
title_full Prediction of High-Risk Cardiac Arrhythmia Based on Optimized Deep Active Learning
title_fullStr Prediction of High-Risk Cardiac Arrhythmia Based on Optimized Deep Active Learning
title_full_unstemmed Prediction of High-Risk Cardiac Arrhythmia Based on Optimized Deep Active Learning
title_short Prediction of High-Risk Cardiac Arrhythmia Based on Optimized Deep Active Learning
title_sort prediction of high risk cardiac arrhythmia based on optimized deep active learning
topic Deep learning
active learning
cardiac arrhythmia
fuzzy logic
url https://ieeexplore.ieee.org/document/10900369/
work_keys_str_mv AT homeyraamiri predictionofhighriskcardiacarrhythmiabasedonoptimizeddeepactivelearning
AT javadmohammadzadeh predictionofhighriskcardiacarrhythmiabasedonoptimizeddeepactivelearning
AT seyedmohsenmirhosseini predictionofhighriskcardiacarrhythmiabasedonoptimizeddeepactivelearning
AT alirezanikravanshalmani predictionofhighriskcardiacarrhythmiabasedonoptimizeddeepactivelearning