Modifications and deviations of electrode positioning for the standard clinical 12 lead ECG is commonplace clinically, primarily in the precordial leads. While multiple clinical studies have aimed to assess the influence such deviations may have on clinical diagnosis and treatment due to resulting variations in ECG morphology, no systematic in-silico study has yet been conducted to the best of our knowledge. We therefore aim to systemically investigate the influence of electrode placement on the morphology of the 12 lead ECG using a cardiac model of electrophysiology during sinus rhythm. Automated perturbation of all electrodes from the subject baseline configuration was performed using a universal torso coordinate (UTC) system for un-attended navigation on the torso surface.
A model for a single healthy subject was previously generated from cardiac MR images and retrofitted with UTCs encoding rotational, longitudinal, and radial directionality within the torso. A baseline configuration for the clinical 12 lead ECG for the single subject was defined by recovering positions of MRI-compatible electrodes within the MRI. Linearly-spaced automatic perturbations were made along the rotational and longitudinal coordinate in the UTCs on the torso surface for each electrode. Bounds were set based on both clinical and anatomical considerations. For each variation in electrode placement, the clinical 12 lead ECG was computed using an efficient reaction-Eikonal model paired with lead-field forward projection from the same prescribed ventricular activation. Changes in the QRS and T-wave due to electrode displacement were quantified in terms of both clinically-relevant ECG metrics and morphology.
The precordial leads that lie in closer proximity to the heart, primarily V1 to V3, experienced the largest morphological changes resulting from electrode perturbation in agreement with previous findings. Little to no deviation was observed in the limb and Goldberger leads.