AI Article Synopsis
- Current ECG diagnostic systems provide classifications but often lack explanations, limiting their clinical use due to the need for manual labeling of extensive data.
- The study introduces a new framework called MA-MIL, which features a multi-layer structure to evaluate ECG data using both a public and private dataset.
- Results showed MA-MIL achieved high accuracy in ECG classification and abnormal segment detection, outperforming existing methods by significant margins, thus enhancing reliability for clinical applications.
Article Abstract
Background And Objective: Current automatic electrocardiogram (ECG) diagnostic systems could provide classification outcomes but often lack explanations for these results. This limitation hampers their application in clinical diagnoses. Previous supervised learning could not highlight abnormal segmentation output accurately enough for clinical application without manual labeling of large ECG datasets.
Method: In this study, we present a multi-instance learning framework called MA-MIL, which has designed a multi-layer and multi-instance structure that is aggregated step by step at different scales. We evaluated our method using the public MIT-BIH dataset and our private dataset.
Results: The results show that our model performed well in both ECG classification output and heartbeat level, sub-heartbeat level abnormal segment detection, with accuracy and F1 scores of 0.987 and 0.986 for ECG classification and 0.968 and 0.949 for heartbeat level abnormal detection, respectively. Compared to visualization methods, the IoU values of MA-MIL improved by at least 17 % and at most 31 % across all categories.
Conclusions: MA-MIL could accurately locate the abnormal ECG segment, offering more trustworthy results for clinical application.
Download full-text PDF |
Source |
|---|---|
| http://dx.doi.org/10.1016/j.cmpb.2024.108164 | DOI Listing |
Publication Analysis
Top Keywords
Similar Publications
Hidden Rhythms: A Complex Case of Hyperemesis Gravidarum-Induced Arrhythmias.
Cureus
December 2024
Obstetrics and Gynecology, Cape Fear Valley Medical Center, Fayetteville, USA.
Hyperemesis gravidarum (HG) is a severe condition marked by intense nausea and vomiting during pregnancy, which is different from typical morning sickness. It is marked by weight loss exceeding 5% of pre-pregnancy weight, ketonuria, dehydration, electrolyte imbalances, and in some cases, arrhythmias - primarily linked to electrolyte disturbances. Treatment typically involves conservative measures such as small, bland meals, medications like metoclopramide and ondansetron, and correction of electrolyte abnormalities.
View Article and Find Full Text PDFAbnormal Exercise Electrocardiography With Normal Stress Echocardiography Is Associated With Subclinical Coronary Atherosclerosis.
Circ Cardiovasc Imaging
January 2025
Duke Clinical Research Institute, Duke University School of Medicine, Durham, NC. (P.S., C.H., G.B., A.H., S.H.S., P.S.D., M.A.D.).
Background: Patients with abnormal (positive) exercise electrocardiography, but normal stress echocardiography (+ECG/-Echo), have an increased risk of adverse cardiovascular events compared with patients with a normal (negative) ECG and a normal stress Echo (-ECG/-Echo). However, it is unclear if +ECG/-Echo discordance is associated with a greater burden of subclinical coronary atherosclerosis.
Methods: Project Baseline Health Study participants who underwent a stress Echo and coronary artery calcium (CAC) scan were stratified by stress Echo result: -ECG/-Echo or +ECG/-Echo.
Epicardial Substrate Ablation in Symptomatic Brugada Syndrome Patients: An Updated Systematic Review and Single-arm Meta-analysis.
Heart Rhythm
January 2025
Dante Pazzanese Institute of Cardiology, Department of Electrophysiology, São Paulo, Brazil.
Background: Brugada syndrome (BrS) is a genetic heart disease that predisposes individuals to ventricular arrhythmias and sudden cardiac death. Although implantable cardioverter-defibrillators (ICDs) and quinidine are primary treatments, recurrent BrS-triggered ventricular arrhythmias can persist. In this setting, epicardial substrate ablation has emerged as a promising alternative for symptomatic patients.
View Article and Find Full Text PDFIdentification of cardiac wall motion abnormalities in diverse populations by deep learning of the electrocardiogram.
NPJ Digit Med
January 2025
Department of Medicine, Stanford University School of Medicine, Stanford, CA, USA.
Cardiac wall motion abnormalities (WMA) are strong predictors of mortality, but current screening methods using Q waves from electrocardiograms (ECGs) have limited accuracy and vary across racial and ethnic groups. This study aimed to identify novel ECG features using deep learning to enhance WMA detection, referencing echocardiography as the gold standard. We collected ECG and echocardiogram data from 35,210 patients in California and labeled WMA using unstructured language parsing of echocardiographic reports.
View Article and Find Full Text PDFA multicenter retrospective study of early cardiac toxicity in operable breast cancer patients receiving concurrent dual or mono anti-HER2 therapy with postoperative radiation therapy.
Breast
January 2025
Department of Radiation Oncology, Ruijin Hospital, Shanghai Jiaotong University School of Medicine, Shanghai, 200025, China; Shanghai Key Laboratory of Proton Therapy, Shanghai, 201801, China. Electronic address:
Purpose: This study aims to assess whether dual anti-HER2 therapy with trastuzumab and pertuzumab increases early cardiac toxicity compared to trastuzumab alone in breast cancer (BC) patients receiving postoperative radiation therapy (RT).
Methods: Consecutive operable BC patients receiving postoperative RT and trastuzumab with or without pertuzumab between January 2017 and September 2020 at seven tertiary hospitals in China were retrospectively reviewed. Cardiac examinations included echocardiography, electrocardiogram (ECG), NT-proBNP, and cTnI at baseline before RT and during the follow-up.
Want AI Summaries of new PubMed Abstracts delivered to your In-box?
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!