College
Rowan-Virtua School of Osteopathic Medicine
Keywords
Artificial Intelligence, Electrocardiogram, Dyspnea, Heart Failure, Emergency Medicine, Diagnostic Accuracy, Machine Learning, Cardiac Dysfunction, NT-proBNP, Clinical Decision Support
Date of Presentation
5-1-2025 12:00 AM
Poster Abstract
Dyspnea is a frequent and diagnostically challenging presentation in emergency departments, requiring rapid distinction between cardiac and non-cardiac causes. Traditional approaches, including NT-proBNP testing and clinical assessment, often lack optimal sensitivity and specificity. Recent advancements in artificial intelligence (AI) have introduced AI-enhanced electrocardiogram (ECG) analysis as a promising diagnostic tool capable of identifying cardiac dysfunction with high accuracy.
A structured literature review was performed using PubMed to identify randomized controlled trials, prospective validations, and systematic reviews published between 2018 and 2024. Studies evaluating AI-ECG performance, NT-proBNP testing, and clinical assessment in adult patients presenting with dyspnea were included. Data extracted included sensitivity, specificity, and area under the receiver operating characteristic curve (AUC) for each diagnostic method. Studies focusing on external validation of AI models and clinical applicability were prioritized to ensure relevance to real-world emergency department settings.
Disciplines
Cardiology | Cardiovascular Diseases | Diagnosis | Emergency Medicine | Health Information Technology | Medicine and Health Sciences | Pathological Conditions, Signs and Symptoms | Pulmonology
Included in
Cardiology Commons, Cardiovascular Diseases Commons, Diagnosis Commons, Emergency Medicine Commons, Health Information Technology Commons, Pathological Conditions, Signs and Symptoms Commons, Pulmonology Commons
Enhancing Diagnostic Accuracy in Dyspnea: The Role of AI-Enhanced ECG Interpretation
Dyspnea is a frequent and diagnostically challenging presentation in emergency departments, requiring rapid distinction between cardiac and non-cardiac causes. Traditional approaches, including NT-proBNP testing and clinical assessment, often lack optimal sensitivity and specificity. Recent advancements in artificial intelligence (AI) have introduced AI-enhanced electrocardiogram (ECG) analysis as a promising diagnostic tool capable of identifying cardiac dysfunction with high accuracy.
A structured literature review was performed using PubMed to identify randomized controlled trials, prospective validations, and systematic reviews published between 2018 and 2024. Studies evaluating AI-ECG performance, NT-proBNP testing, and clinical assessment in adult patients presenting with dyspnea were included. Data extracted included sensitivity, specificity, and area under the receiver operating characteristic curve (AUC) for each diagnostic method. Studies focusing on external validation of AI models and clinical applicability were prioritized to ensure relevance to real-world emergency department settings.
