A comprehensive framework for evaluation of high pacing frequency and arrhythmic optical mapping signals

Figures

• 

  FIGURE 1. uAT mapping extracted from a fixed time window during reentry. (A) Background optical mapping image of the endocardial surface from a coronary-perfused right ventricle. The region of interest lying in between two line electrodes for reentry induction is highlighted by blue shading. (B) Example pre-processed optical mapping signals extracted from the full region of interest. (C) A snapshot of the dynamic voltage-dependent optical signal (F/dF) during reentry [time indicated by the red dashed line in (B)]. (D) An instantaneous phase map derived using the Hilbert transform showing evidence for a repetitive propagating wave front [red dashed line in (B)]. (E) Example pre-processed optical mapping signals extracted from pixels along the circular black arrow in (D). A green shaded region indicates a uniform time window used to derive uAT points using the uniform window method for calculation of activation time. The red arrow shows the principle sequence of activation. (F) uAT map corresponding to the fixed time window shown in B during reentry. Isolines are spaced 10 ms.

• 

  FIGURE 2. Schematic workflow of the novel pAT mapping framework.

• 

  FIGURE 3. Defining phase window-independent upstroke morphology and pAT using the signal derivative. (A) Extraction of the pre-processed optical action potential signal during reentry from pixel coordinate (20,10). Corresponding phase windows of pAT are superimposed. The derivative of the corresponding signal is shown. Pronounced derivative peaks falling within phase windows indicate pAT (red lines). (B) Similarly, an optical action potential signal and corresponding derivative extracted from an in silico model of reentry. (C) Membrane voltage (Vm) map extracted from the epicardial surface of the electrical in silico model at the time indicated by a black arrow in (B). (D) Corresponding F/dF computed from the in silico model. In silico optical signal in B extracted from pixel indicated by +.

• 

  FIGURE 4. Adapted connected components algorithm for wave front clustering. (A) A schematic diagram illustrates pAT points on a space-time plot. In the example, wave front label assignment has been made for all pAT points in the first 4 frames to label #1. Let the current pAT point be evaluated for connectivity be at pixel 4, frame 4 on the space-time plot. The red box illustrates the space-time bounds of eligible connectivity. In this case, the currently unassigned pAT point at pixel 5, frame 7 lies within the connectivity boundary and will be assigned to wave front label #1. (B) A spatio-temporal distribution of pAT points for the first 4 s of pacing and reentry induction in experiments (upper panel). pAT points are assigned colors based on their wave front label assignment (lower panel). (C) Similarly, pAT points of the in silico model from the moment of reentry induction (upper panel) and wave front label assignment (lower panel).

• 

  FIGURE 5. Spatial classification of activation sources. (A) A dynamic pAT map (left panel). (B) Origins of activation for breakthrough (red triangles) and passive (blue triangles) sources. (C) Leading pAT points of the active wave front (pink squares). (D) Conduction vectors oriented in the direction of local propagation overlaid on (C). (E) The conduction fields of leading pAT points. (F) Number of accumulated breakthrough or passive origins throughout wave front #4. (G) A pAT map shows leading wave front pAT sites (purple squares) with conduction fields emanating from a single breakthrough (red triangles) and passively from outside of the imaged field of view. (H) The leading wave front during tachyarrhythmia was also maintained by reentrant propagation through myocardial pathways or regions of tortuous propagation. Yellow arrows indicate the local direction of propagation.

• 

  FIGURE 6. Dynamic mapping of experimental complex activity. (A) A spatio-temporal distribution of pAT points spanning pacing, induction and reentry phases of the experimental optical acquisition. pAT maps were dynamically probed for wave front labels (#): 1 (B), 3 (C), 5 (D–F), 6 (G–K), and 7 (L). Lines of conduction slowing where the direction of maximal gradient exceedes 50 ms are annotated (pink lines). pAT windows are labelled in panel (A) (red). (M) Lines of conduction slowing superimposed. Lines are colored according to their average AT.

• 

  FIGURE 7. Repetitive/non-repetitive classification of wave fronts. (A—D) Wave fronts shown in Figure 5 were classified as repetitive or non-repetitive. The spatial distribution of the number of repetitions of activation for each wave front (upper panels) and their histograms (lower panels) are shown. Labels were classified as repetitive if the histogram peak of activation repetitions was ≥1.

• 

  FIGURE 8. Simulations of pAT mapping and image resolution. (A) Ground truth eAT derived from a simulation of the underlying electrical action potential. (B) Optical mapping signals were derived from electrical simulations in (A). The effects of image resolution evaluated on uAT and pAT derived from optical mapping signals. Optical mapping signals were derived from downsampled electrical simulations to simulate reduction of image resolution from 400 μm to 1,000 µm. (C) Mean ± standard deviation eAT, uAT, and pAT from dynamic AT maps of the full simulation. (D) Linear regression of pixel-to-pixel correlations of eAT with uAT and pAT.

• 

  FIGURE 9. Influences of the signal-to-noise ratio. (A) Action potential traces with varying lowpass filter cut-off frequencies. (B) Mean ± standard deviation signal-to-noise ratios for experimental optical mapping recordings during S1S2S3S4 stimulation and tachyarrhythmia induction (N = 5). (C) Spatio-temporal distribution of pAT points corresponding to recordings used in A.

• 

  FIGURE 10. pAT mapping tachyarrhythmia in chronic myocardial infarction. (A) The endocardial surface was subject to optical mapping, encompassing a region of scar in the ventricular apex. (B) Background optical mapping signals. (C) Optical mapping traces taken from normal myocardium [black asterisk in (B)] and the scar region [red asterisk in (B)]. (D) Map of SNR from the same optical mapping recording. (E) pAT matrix with labelled wave fronts and indicating the timing stimulation pulses (red arrows). (F–M) Progressive pAT maps during the onset of tachyarrhythmia.