Noninvasive identification of two lesions with local repolarization changes using two dipoles in inverse solution...

Area of Science:

• Computational Electrophysiology
• Medical Imaging & Diagnostics
• Cardiovascular Research

Background:

• Accurate localization of cardiac electrical abnormalities is crucial for diagnosing ventricular conditions.
• Simultaneous repolarization changes in the ventricular myocardium, often indicative of two-vessel disease, present a diagnostic challenge.

Purpose of the Study:

• To propose and validate a novel method for the inverse localization and identification of two distinct, simultaneous ventricular repolarization lesions.
• To assess the robustness of this method against errors and noise in the input body surface potential map data.

Main Methods:

• Calculated inverse solutions from the difference between STT integral body surface potential maps with and without repolarization changes.
• Modeled two simultaneous lesions using a two-dipole representation in a numerical ventricular model, introducing various noise types to input data.
• Employed discriminant analysis with feature selection to distinguish correct inverse solutions identifying two lesions from incorrect ones.

Main Results:

• Achieved a mean localization error of 1.1±0.7cm for identified double lesions.
• Demonstrated high sensitivity and specificity exceeding 90% using quadratic discriminant analysis with cross-validation and feature selection.

Conclusions:

• The combined approach of inverse solutions with a two-dipole model and discriminant analysis effectively identifies two simultaneous ventricular lesions.
• This method successfully identifies multiple lesions without requiring prior knowledge of their number, enhancing diagnostic capabilities.