Analytical approaches for myocardial fibrillation signals

Abstract

Atrial and ventricular fibrillation are complex arrhythmias, and their underlying mechanisms remain widely debated and incompletely understood. This is partly because the electrical signals recorded during myocardial fibrillation are themselves complex and difficult to interpret with simple analytical tools. There are currently a number of analytical approaches to handle fibrillation data. Some of these techniques focus on mapping putative drivers of myocardial fibrillation, such as dominant frequency, organizational index, Shannon entropy and phase mapping. Other techniques focus on mapping the underlying myocardial substrate sustaining fibrillation, such as voltage mapping and complex fractionated electrogram mapping. In this review, we discuss these techniques, their application and their limitations, with reference to our experimental and clinical data. We also describe novel tools including a new algorithm to map microreentrant circuits sustaining fibrillation.

Keywords: Analysis; Atrial fibrillation; Complex fractionated electrograms; Dominant frequency; Organizational index; Phase analysis; Rotors; Shannon entropy; Ventricular fibrillation; Voltage mapping.

Fig. 1

Dominant Frequency Maps of VF: (A) Optical action potential data for two pixels during a recording of ventricular fibrillation in a langendorff perfused rat heart. (B) Dominant frequency map: frequency spectrum of both these pixels is shown (left) and the frequency with the largest power for each pixel is plotted on a colour map. The corresponding histogram of DF values for the entire heart is also shown (bottom right). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 2

Shannon Entropy Maps of VF: The optical action potential data for the two pixels during a recording of ventricular fibrillation shown in Fig. 1 was used to construct a Shannon entropy map. The fluorescence signal from optical mapping of each pixel is placed in fluorescence amplitude bins with frequency of occurrence over the recording period on the y axis (left), and the Shannon entropy of each pixel was calculated using the relative probability of the signal falling in each fluorescence amplitude bin. The Shannon entropy values for each pixel are then depicted on a Shannon entropy map (top right), the distribution of SE values shown on the histogram (bottom right).

Fig. 3

Calculating phase in optically mapped VF: (A–F) Steps for calculating the phase are as follow. (A) Tag minima (green circles) of the signal using a sliding window, then tag maxima (red circles) between each pair of minima, and finally remove any small amplitude pairs of maxima-minima. (B) Cubic spline fits were performed on the maxima and minima of the signal, and the average of these maxima (red) and minima (green) splines (the mean line, shown in black) was subtracted from the signal to give a signal of zero mean. (C) The real and imaginary parts of the Hilbert transform of this zero-mean signal were plotted in the phase plane. (D) The angle around the trajectory in (C) gives the phase angle. A phase map of ventricular fibrillation at a single time point is shown in (E), and a phase singularity heat map can be used to show sites with high incidences phase singularities over time (F). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 4

Outline of filtering and processing steps involved in phase calculation of unipolar and bipolar electrogram signals. Raw unipolar electrograms are pre-processed (a–c). Both unipolar electrogram derivatives and bipolar electrograms are then pre-processed (d–f) before tagging of individual activations (g–h). Phase is then calculated identically for both modalities of signal (i–n) to give unipolar and bipolar phase (o–p). [Figure reproduced from Roney et al. [46], in line with Creative Commons Attribution license (CC BY 4.0)].

Fig. 5

Left atrial 3D reconstruction showing phase singularity heat maps during AF. Regions of high and low phase singularity during AF are shown for three patients, with maps created using bipolar electrogram phase analysis. [Figure reproduced from Roney et al. [46], in line with Creative Commons Attribution license (CC BY 4.0)].

Fig. 6

AF voltage maps correlate well with MRI scar maps. Electroanatomical voltage maps in AF (left), with their corresponding electrograms. The AF voltage maps were created using 8 s segments of recordings, and the maps represents the mean peak-to-peak AF voltage (AF-V) (on a research version of EnsiteTM Velocity, SJM). The LGE-CMRI derived atrial scar maps (right) are imported, and registered to the left atrial geometries. This representative example illustrates good correlation between regions of low mean AF voltage and scar as defined by LGE-CMRI.

Fig. 7

Processing and analysis of conduction block at single-cell level in HL1-6 myocyte monolayer. (A) Processing stage of algorithm: each pixel in the activation map is iterated through (first graphic). For each pixel, its neighbourhood is determined and then the absolute activation time difference between it and each neighbouring pixel determined (next two graphics). After repeating for all neighbourhood pixels for every pixel in the activation map, a heatmap highlighting areas of large activation differences is generated (last two graphics). (B) This process can be repeated for multiple activation maps generated from the same recording. When these heatmaps are summed and averaged, the resulting heatmap isolates the ‘constant’ features of the rotational activity. (C) The final heatmap can be overlaid on wheat germ agglutinin membrane staining to study the correlation between cell morphology, core shape and/or core location.