Related Experiment Video
Updated: Oct 15, 2025
06:07
Analyzing Long-Term Electrocardiography Recordings to Detect Arrhythmias in Mice
Published on: May 23, 2021
4.0K
Machine learning-derived electrocardiographic algorithm for the detection of cardiac amyloidosis
Lore Schrutka1, Philip Anner1,2, Asan Agibetov2
1Department of Internal Medicine II, Division of Cardiology, Medical University of Vienna, Vienna, Austria.
Heart (British Cardiac Society)
|October 30, 2021
Summary
A new ECG-based tool aids in diagnosing cardiac amyloidosis (CA) by analyzing electrophysiological patterns. This machine learning approach improves CA detection without advanced imaging.
Area of Science:
- Cardiology
- Medical Imaging
- Machine Learning
Background:
- Cardiac amyloidosis (CA) diagnosis relies on advanced imaging, as surface ECG patterns have limited diagnostic value.
- Current diagnostic methods for CA present challenges due to their complexity and invasiveness.
Purpose of the Study:
- To conduct a detailed electrophysiological characterization of patients with CA.
- To develop a user-friendly diagnostic tool for CA based on surface ECG.
Main Methods:
- Electrocardiographic imaging (ECGI) was used to create electroanatomical maps in CA patients and controls.
- A machine learning approach analyzed ECGI data to generate a surface ECG-based diagnostic tool.
- Correlations were drawn between ECGI findings and standard 12-lead ECG recordings.
Main Results:
- Low voltage areas and specific activation patterns were identified in CA patients' ventricles via ECGI.
- Surface ECG patterns correlated with ventricular activation and epicardial voltage.
- A diagnostic tool trained on these patterns significantly improved CA detection rates by cardiologists (AUC 0.97).
Conclusions:
- A machine learning-based ECG algorithm was developed using electroanatomical mapping data from CA patients.
- This simple ECG tool can help suspect CA, reducing reliance on advanced imaging techniques.
