Selvester scoring in patients with strict LBBB using the QUARESS software

Area of Science:

• Cardiology
• Medical Imaging
• Computational Biology

Background:

• The Selvester scoring method estimates infarct size from body-surface ECGs in post-myocardial infarction (MI) patients.
• Automating this score aims to increase measurement speed and reduce inter-observer variability, which currently requires significant electrocardiography expertise.

Purpose of the Study:

• To evaluate the accuracy and quality of the QuAReSS software for automated Selvester scoring.
• To assess the software's performance in a set of standard 12-lead ECGs from post-MI patients.

Main Methods:

• Standard 12-lead ECGs were analyzed from 67 post-MI patients with strict left bundle branch block (LBBB) patterns.
• The QuAReSS software was used for semi-automatic Selvester score measurement by two independent cardiologist groups.
• Agreement metrics were employed to compare scores between groups and between manual and automatic measurements.

Main Results:

• The average absolute difference in Selvester scores between the two manual groups was 1.4±1.5 points.
• Differences between the automatic QuAReSS method and manual adjudications were minimal (1.2±1.2 and 1.3±1.2 points).
• Score agreement within two points was 82% for manual measurements and 84-90% for automatic vs. manual comparisons.

Conclusions:

• The QuAReSS software provides valid Selvester score measurements in patients with strict LBBB.
• The software requires minimal correction from cardiologists, indicating its reliability for automated infarct size estimation.