Detection of proximal coronary occlusion in acute coronary syndrome: a feasibility study using computerized
Markku J Eskola1, Kjell C Nikus, Liisa-Maria Voipio-Pulkki
1Heart Center, Tampere University Hospital, Tampere, Finland. markku.eskola@pshp.fi
Insights
A new computer model accurately detects proximal left anterior descending (LAD) artery occlusion using electrocardiograms (ECG). This tool aids in rapid myocardial infarction (MI) diagnosis, improving patient care.
Area of Science:
- Cardiology
- Medical Imaging
- Artificial Intelligence in Medicine
Background:
- Rapid identification of coronary artery occlusion is critical for myocardial infarction (MI) care.
- Electrocardiograms (ECGs) can identify occlusions with expert interpretation.
- Computer-based ECG analysis requires validation against manual methods.
Purpose of the Study:
- Develop a computer-assisted model for detecting proximal left anterior descending (LAD) coronary artery occlusion.
- Validate the performance of the computer model against manual ECG interpretation.
Main Methods:
- Constructed a computerized ECG model based on manual anatomical interpretation from 216 suspected acute coronary syndrome (ACS) patients.
- Determined agreement between cardiologists' manual ECG evaluation and the computerized algorithm for LAD occlusion detection.
Main Results:
- The computer model achieved 100% specificity and 86% sensitivity for proximal LAD occlusion (kappa=0.72).
- Overall LAD occlusion detection showed 99% specificity and 67% sensitivity (kappa=0.71).
- Distal LAD occlusion detection had 99% specificity and 40% sensitivity (kappa=0.72).
Conclusions:
- Computerized anatomical interpretation of ECGs is feasible for detecting LAD occlusion.
- The developed model demonstrates excellent accuracy in identifying proximal LAD occlusions.
Background:
Rapid identification of a proximal occlusion site of a major coronary artery is of paramount importance in the care of myocardial infarction (MI). It is increasingly recognized that routine electrocardiogram (ECG) can be used for that purpose, provided that expert interpretation is available. Computer-based signal analysis has potential to enhance early ECG interpretation but its performance must be validated against manual algorithms. We therefore set out to develop a computer-assisted model to detect proximal occlusion of the left anterior descending coronary artery (LAD) in patients with suspected acute coronary syndrome (ACS).
Methods:
Based on manual anatomical interpretation of the ECG, obtained from 216 consecutive patients who were admitted due to suspected ACS, an automatic computerized ECG model to detect LAD occlusion was constructed. Agreement between manual evaluation of the ECG by two cardiologists and a computerized ECG algorithm to detect occlusion of the LAD and the site of occlusion was determined.
Results:
Using an expert electrocardiographer's anatomical interpretation as the gold standard, the computer model recognized patients fulfilling ECG criteria for any occlusion of the LAD with a specificity of 99% and a sensitivity of 67% (kappa= 0.71). However, proximal LAD occlusion was detected with 100% specificity and 86% sensitivity (kappa= 0.72). The computer program detected a distal occlusion in the LAD with a specificity of 99% and a sensitivity of 40% (kappa= 0.72).
Conclusions:
Computerized anatomical interpretation of the ECG is feasible and allows detection of a proximal LAD occlusion with excellent accuracy.
Related Concept Videos
01:30Acute Coronary Syndrome III: Diagnostic Studies
01:30Acute Coronary Syndrome I: Introduction
01:29Electrocardiogram
Three major waveforms are present in a typical ECG recording: the P wave, the QRS complex, and the T...
01:19Acute Coronary Syndrome II: Pathophysiology and Clinical Manifestations
01:25Imaging Studies for Cardiovascular System VI: Calcium -Scoring CT
01:28Imaging Studies for Cardiovascular System V: CT