Feasibility and Accuracy of Cardiac Magnetic Resonance Imaging-Based Whole-Heart Inverse Potential Mapping of Sinus Rhythm and Idiopathic Ventricular Foci

How Accurate Is Inverse Electrocardiographic Mapping? A Systematic In Vivo Evaluation

BACKGROUND

Inverse electrocardiographic mapping reconstructs cardiac electrical activity from recorded body surface potentials. This noninvasive technique has been used to identify potential ablation targets. Despite this, there has been little systematic evaluation of its reliability.

METHODS

Torso and ventricular epicardial potentials were recorded simultaneously in anesthetized, closed-chest pigs (n=5), during sinus rhythm, epicardial, and endocardial ventricular pacing (70 records in total). Body surface and cardiac electrode positions were determined and registered using magnetic resonance imaging. Epicardial potentials were reconstructed during ventricular activation using experiment-specific magnetic resonance imaging-based thorax models, with homogeneous or inhomogeneous (lungs, skeletal muscle, fat) electrical properties. Coupled finite/boundary element methods and a meshless approach based on the method of fundamental solutions were compared. Inverse mapping underestimated epicardial potentials >2-fold (P<0.0001).

RESULTS

Mean correlation coefficients for reconstructed epicardial potential distributions ranged from 0.60±0.08 to 0.64±0.07 across all methods. Epicardial electrograms were recovered with reasonable fidelity at ≈50% of sites (median correlation coefficient, 0.69-0.72), but variation was substantial. General activation spread was reproduced (median correlation coefficient, 0.72-0.78 for activation time maps after spatio-temporal smoothing). Epicardial foci were identified with a median location error ≈16 mm (interquartile range, 9-29 mm). Inverse mapping with meshless method of fundamental solutions was better than with finite/boundary element methods, and the latter were not improved by inclusion of inhomogeneous torso electrical properties.

CONCLUSIONS

Inverse potential mapping provides useful information on the origin and spread of epicardial activation. However the spatio-temporal variability of recovered electrograms limit resolution and must constrain the accuracy with which arrhythmia circuits can be identified independently using this approach.

Novel algorithm for accelerated electroanatomic mapping and prediction of earliest activation of focal cardiac arrhythmias using mathematical optimization

Background

Premature beats (PBs) are a common finding in patients suffering from structural heart disease, but they can also be present in healthy individuals. Catheter ablation represents a suitable therapeutic approach. However, the exact localization of the origin can be challenging, especially in cases of low PB burden during the procedure.

Objective

The aim of this study was to develop an automated mapping algorithm on the basis of the hypothesis that mathematical optimization would significantly accelerate the localization of earliest activation.

Methods

The algorithm is based on iterative regression analyses. When acquiring local activation times (LATs) within a 3-dimensional anatomic map of the corresponding heart chamber, this algorithm is able to identify that exact position where a next LAT measurement adds maximum information about the predicted site of origin. Furthermore, on the basis of the acquired LAT measurements, the algorithm is able to predict earliest activation with high accuracy.

Results

A systematic retrospective analysis of the mapping performance comparing the operator with simulated search processes by the algorithm within 17 electroanatomic maps of focal spreading arrhythmias revealed a highly significant reduction of necessary LAT measurements from 55 ± 8.8 to 10 ± 0.51 (n = 17; P < .0001).

Conclusion

On the basis of mathematical optimization, we developed an algorithm that is able to reduce the number of LAT measurements necessary to locate the site of earliest activation. This algorithm might significantly accelerate the mapping procedure by guiding the operator to the optimal position for the next LAT measurement. Furthermore, the algorithm would be able to predict the site of origin with high accuracy early during the mapping procedure.

Integrated whole-heart computational workflow for inverse potential mapping and personalized simulations

Background

Integration of whole-heart activation simulations and inverse potential mapping (IPM) could benefit the guidance and planning of electrophysiological procedures. Routine clinical application requires a fast and adaptable workflow. These requirements limit clinical translation of existing simulation models. This study proposes a comprehensive finite element model (FEM) based whole-heart computational workflow suitable for IPM and simulations.

Methods

Three volunteers and eight patients with premature ventricular contractions underwent body surface potential (BSP) acquisition followed by a cardiac MRI (CMR) scan. The cardiac volumes were segmented from the CMR images using custom written software. The feasibility to integrate tissue-characteristics was assessed by generating meshes with virtual edema and scar. Isochronal activation maps were constructed by identifying the fastest route through the cardiac volume using the Möller–Trumbore and Floyd–Warshall algorithms. IPM’s were reconstructed from the BSP’s.

Results

Whole-heart computational meshes were generated within seconds. The first point of atrial activation on IPM was located near the crista terminalis of the superior vena cave into the right atrium. The IPM demonstrated the ventricular epicardial breakthrough at the attachment of the moderator band with the right ventricular free wall. Simulations of sinus rhythm were successfully performed. The conduction through the virtual edema and scar meshes demonstrated delayed activation or a complete conductional block respectively.

Conclusion

The proposed FEM based whole-heart computational workflow offers an integrated platform for cardiac electrical assessment using simulations and IPM. This workflow can incorporate patient-specific electrical parameters, perform whole-heart cardiac activation simulations and accurately reconstruct cardiac activation sequences from BSP’s.

Electronic supplementary material

The online version of this article (doi:10.1186/s12967-016-0902-0) contains supplementary material, which is available to authorized users.

Feasibility and Accuracy of Cardiac Magnetic Resonance Imaging-Based Whole-Heart Inverse Potential Mapping of Sinus Rhythm and Idiopathic Ventricular Foci

Background

Inverse potential mapping (IPM) noninvasively reconstructs cardiac surface potentials using body surface potentials. This requires a volume conductor model (VCM), usually constructed from computed tomography; however, computed tomography exposes the patient to harmful radiation and lacks information about tissue structure. Magnetic resonance imaging (MRI) is not associated with this limitation and might have advantages for mapping purposes. This feasibility study investigated a magnetic resonance imaging–based IPM approach. In addition, the impact of incorporating the lungs and their particular resistivity values was explored.

Methods and Results

Three volunteers and 8 patients with premature ventricular contractions scheduled for ablation underwent 65‐electrode body surface potential mapping. A VCM was created using magnetic resonance imaging. Cardiac surface potentials were estimated from body surface potentials and used to determine the origin of electrical activation. The IPM‐defined origin of sinus rhythm corresponded well with the anatomic position of the sinus node, as described in the literature. In patients, the IPM‐derived premature ventricular contraction focus was 3‐dimensionally located within 8.3±2.7 mm of the invasively determined focus using electroanatomic mapping. The impact of lungs on the IPM was investigated using homogeneous and inhomogeneous VCMs. The inhomogeneous VCM, incorporating lung‐specific conductivity, provided more accurate results compared with the homogeneous VCM (8.3±2.7 and 10.3±3.1 mm, respectively; P=0.043). The interobserver agreement was high for homogeneous (intraclass correlation coefficient 0.862, P=0.003) and inhomogeneous (intraclass correlation coefficient 0.812, P=0.004) VCMs.

Conclusion

Magnetic resonance imaging–based whole‐heart IPM enables accurate spatial localization of sinus rhythm and premature ventricular contractions comparable to electroanatomic mapping. An inhomogeneous VCM improved IPM accuracy.