Search research articles

Related Concept Videos

Imaging Studies for Cardiovascular System I:Echocardiography

Imaging Studies for Cardiovascular System I:Echocardiography

Cardiac imaging studies encompass a wide range of noninvasive and minimally invasive techniques designed to visualize the heart's structure and function in detail. One such technique is echocardiography, which uses high-frequency ultrasound waves to produce detailed images of the heart, known as echocardiograms.
Indications: Echocardiography is utilized to diagnose heart failure, valve disorders, and myocardial infarction. It also assesses cardiac structures' size, shape, and motion, evaluates...

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Home
Performance Of Off-the-shelf Machine Learning Architectures And Biases In Low Left Ventricular Ejection Fraction Detection.

Home
Performance Of Off-the-shelf Machine Learning Architectures And Biases In Low Left Ventricular Ejection Fraction Detection.

Related Experiment Video

Ultrasonic Assessment of Myocardial Microstructure

Ultrasonic Assessment of Myocardial Microstructure

Published on: January 14, 2014

Performance of off-the-shelf machine learning architectures and biases in low left ventricular ejection fraction

Jake A Bergquist^1,2,3, Brian Zenger⁴, James Brundage⁴

¹Scientific Computing and Imaging Institute, University of Utah, Salt Lake City, Utah.

Heart Rhythm O2

|November 4, 2024

View abstract on PubMed

Summary

This summary is machine-generated.

Off-the-shelf artificial intelligence-machine learning (AI-ML) models can effectively detect low left ventricular ejection fraction (LVEF) from electrocardiograms (ECGs). However, patient characteristics like race and sex may impact AI-ML prediction accuracy, highlighting potential biases.

Keywords:

Artificial intelligence Electrocardiogram Explainability Heart failure Machine learning

More Related Videos

Murine Echocardiography of Left Atrium, Aorta, and Pulmonary Artery

Murine Echocardiography of Left Atrium, Aorta, and Pulmonary Artery

Published on: February 20, 2017

Evaluation of Left Ventricular Structure and Function using 3D Echocardiography

Evaluation of Left Ventricular Structure and Function using 3D Echocardiography

Published on: October 28, 2020

Related Experiment Videos

Ultrasonic Assessment of Myocardial Microstructure

Ultrasonic Assessment of Myocardial Microstructure

Published on: January 14, 2014

Murine Echocardiography of Left Atrium, Aorta, and Pulmonary Artery

Murine Echocardiography of Left Atrium, Aorta, and Pulmonary Artery

Published on: February 20, 2017

Evaluation of Left Ventricular Structure and Function using 3D Echocardiography

Evaluation of Left Ventricular Structure and Function using 3D Echocardiography

Published on: October 28, 2020

Area of Science:

Cardiology
Artificial Intelligence
Machine Learning
Medical Informatics

Background:

Artificial intelligence-machine learning (AI-ML) offers novel methods for extracting clinical insights from electrocardiograms (ECGs).
Numerous open-source AI-ML architectures exist, adaptable for various applications.
Limited research has explored the utility and limitations of these "off-the-shelf" AI-ML models in ECG analysis.

Purpose of the Study:

To evaluate the effectiveness of readily available AI-ML architectures for ECG analysis.
To identify which "off-the-shelf" AI-ML models are suitable for ECG interpretation.
To understand the failure modes and potential biases of AI-ML approaches in ECG-based LVEF detection.

Main Methods:

Applied six "off-the-shelf" AI-ML architectures to a large cohort of 24,868 ECGs.

Assessed the performance of these models in detecting low left ventricular ejection fraction (LVEF).

Investigated patient characteristics associated with inaccurate LVEF predictions (false positives/negatives).

Main Results:

All tested architectures achieved an area under the receiver-operating characteristic curve (AUC) above 0.9 for LVEF detection.
The ResNet 18 architecture demonstrated the highest performance with an average AUC of 0.917.
Patient factors including race, sex, and comorbidities were linked to reduced LVEF prediction accuracy.

Conclusions:

"Off-the-shelf" AI-ML architectures can achieve performance comparable to custom-built models for ECG analysis.
The study identified potential biases related to patient characteristics in AI-ML ECG interpretation.
Findings emphasize the need for careful consideration of efficiency and equity in AI-ML deployment for healthcare.