A robust conduction velocity pipeline method for guiding evaluation of the electrophysiological effect due to radiofrequency catheter ablation: outcomes from an isolated mice atria model
Introduction: atrial fibrillation (AF) is the most common sustained cardiac arrhythmia in clinical practice, affecting 1-2% of the general population. The treatment of AF by antiarrhythmic drugs are of limited effectiveness, and radiofrequency catheter ablation (RFA) is the most effective treatment to terminate and prevent AF recurrence. AF non-termination or its recurrence, after RFA, can occur when the cauterized atrium tissue still conductive, or with its conduction's recovery. Different parameters during the RFA procedure influence durable electrical inactivity of the cells (tissue thickness, electrode-tissue interface temperature, electrode size, electrode-tissue contact pressure, and duration of RFA delivery). Through an experimental isolated mice atria model, different RFA times were performed, and a robust pipeline method for conduction velocity (CV) estimation is proposed to evaluate the RFA electrophysiological effects and lesion size on the atria tissue.
Methods: the electrical activity of the isolated right atrium of rats (n=2), under different RFA strategies on the epicardium (0.5, 1.5, 2, 2.5, 3, 4 s) applied in the center area was acquired in the endocardium through an Optical Mapping system (OM). Sequences of fluorescence images were acquired at a resolution of 82 x 82 pixels, at 868 Hz. Optical action potentials (OAPs) were filtered with an adaptive 3-D Gaussian low-pass filter. The robustness of local activation detection (LAT) under different RFA times delivery was evaluated with a proposed post-processing pipeline and two distinct LAT detection techniques (first derivative and 50% criteria). Then, 2D surface isochronous maps are calculated to allow estimation of CV, by a proposed method called a Circle Method (CM), allowing identification of the wavefront propagation direction, velocity of propagation, and estimation of the ablated lesion area. Action Potential Duration (APD30, 50, and 90), Dominant Frequency (DF), and Organization Index (OI) were calculated and compared between three distinct areas of the wavefront propagation (earliest activated, most delayed, and inside the ablated area). Moreover, the time and frequency metrics, calculated from electrograms acquired simultaneously in the epicardium are compared with those obtained in the respective endocardium optical recordings.
Results: post-processing improves signal-to-noise ratio enhancing LAT detection. After post-processing filters, we can see lower ∆LAT values, especially after 2.5 s of RFA, and the 50% criteria were the method with less difference, variations, and closer to the pattern (results outside the RFA area). Then, 2D surface isochronous maps are calculated based on 50% criteria and CV estimated with the proposed CM. In the ablated area, the CV decreases as the radius chosen for the circle is closer to the central point. Lower values of CV were found in the ablated area, decreasing from 90 to 40 cm/s, highlighting the electrical inactive areas, and consequently the RFA lesion. The morphology of the optical and electrical signals with prolonged ablation time was altered as the ablation time increased. The APD90, especially in the ablated area, changes from 57 to 117 ms. DF and OI decreased with the increase of the ablation time and resulted in fragmented electrograms.
Conclusion: for low levels of signal-to-noise ratio and inside the ablated areas, the 50% criteria method should be used. The proposed CM allowed robust estimation of wavefront direction, CV, and RFA lesion size. Atria electrophysiological modifications induced by the ablation were seen in traditional metrics calculated from both electrical and optical recordings.