Analytical approaches for myocardial fibrillation signals

Balvinder S. Handa, Caroline H. Roney, Charles Houston, Norman A. Qureshi, Xinyang Li, David S. Pitcher, Rasheda A. Chowdhury, Phang Boon Lim, Emmanuel Dupont, Steven A. Niederer, Chris D. Cantwell, Nicholas S. Peters, Fu Siong NgComputers in Biology and Medicine 2018 · DOI

TL;DR

This paper reviews current analytical techniques for understanding myocardial fibrillation (AF and VF), highlighting their limitations in accurately interpreting complex electrical signals. The authors discuss established methods like dominant frequency and phase mapping, alongside novel approaches developed in their lab for identifying microreentrant circuits, aiming to improve our understanding and treatment of these arrhythmias.

What Problem Does This Paper Address?

Myocardial fibrillation, both atrial (AF) and ventricular (VF), are life-threatening arrhythmias whose underlying mechanisms are still not fully understood. This lack of understanding stems from the inherent complexity of the electrical signals recorded during these events. These signals often appear chaotic, making it difficult to discern organized patterns or identify the specific drivers that sustain the fibrillation. Existing analytical tools struggle to accurately translate these complex signals into interpretable information about the propagating wavefronts and the myocardial substrate involved. This paper aims to provide a comprehensive overview of the analytical approaches used to tackle this challenge, discuss their strengths and weaknesses, and introduce new methods for a deeper understanding.

Methods

The paper categorizes analytical approaches into two main groups: those that map putative drivers of fibrillation and those that map the underlying myocardial substrate.

Mapping Putative Drivers:

  • Dominant Frequency (DF) Analysis: Identifies the frequency with the highest power in a signal's spectrum, hypothesizing that areas with the highest DF drive the fibrillation.
  • Organizational Index (OI): A more stable measure than DF, defined by the ratio of DF power to total spectrum power, proposed to better identify AF drivers.
  • Shannon Entropy (SE): Measures signal unpredictability, with higher SE values potentially indicating areas of complex, fragmented signals at the core of rotational activity.
  • Phase Mapping: A technique to localize phase singularities, believed to be at the core of rotational drivers. This involves processing signals (e.g., optical mapping data, electrograms) to create phase angles and then mapping these across the tissue. The paper details a specific method using the Hilbert transform for phase calculation.

Mapping Myocardial Substrate:

  • Voltage Mapping: Uses electrogram voltage amplitude as a surrogate for scar tissue, which can harbor arrhythmogenic substrates. This is often performed in sinus rhythm or during pacing for better clarity.
  • Complex Fractionated Atrial Electrograms (CFAEs): Characterizes electrograms with complex morphologies, hypothesized to represent areas of conduction block, wavefront collisions, or microreentrant circuits.

Novel Tools:

  • The authors introduce a novel algorithm for identifying lines of conduction block at a single-cell level. This algorithm analyzes activation time disparities between neighboring pixels to generate "heat-maps" that highlight areas of consistent conduction block or slowing, which can then be overlaid on microscopy images to correlate with cellular morphology. This method aims to distinguish between microreentrant circuits and other forms of reentry.

Datasets

The paper discusses the application of these analytical techniques to both experimental (e.g., rat heart models, HL1-6 cell lines) and clinical data (e.g., human atrial fibrillation recordings from multipolar catheters). Specific datasets or benchmarks are not explicitly introduced as new contributions, but rather the paper reviews how existing datasets are analyzed using the discussed methods.

Key Insights and Ideas

  • Complexity is the Enemy: The fundamental challenge is deciphering organized electrical activity from seemingly chaotic signals during fibrillation.
  • Drivers vs. Substrate: Understanding fibrillation requires looking at both the "engines" (drivers) and the "terrain" (substrate) that sustain it.
  • Limitations of Current Tools: DF mapping can be unstable, and while phase mapping is powerful, its clinical application is hampered by resolution and coverage issues of current catheters.
  • Microreentry as a Key Mechanism: There's growing evidence that microreentrant circuits, rather than just single rotors, might be crucial drivers, especially at a cellular level.
  • Novel Algorithm for Conduction Block: The developed algorithm offers a new way to visualize and quantify conduction block at a fine scale, potentially revealing the micro-architectural basis of reentrant circuits. This is a significant step beyond simply identifying phase singularities.
  • Mean AF Voltage as a Potential Marker: The authors suggest that mean AF voltage, when sampled adequately, can be a reproducible marker correlating with scar and reflecting the functional response of the underlying substrate.

Key Results

  • The paper provides illustrative examples of how DF maps and SE maps are generated and interpreted in experimental models (e.g., rat VF).
  • It highlights that while phase mapping can identify phase singularities, its clinical translation for ablation guidance has yielded mixed results, suggesting limitations in current catheter technology.
  • The authors' novel algorithm for identifying conduction block has demonstrated that the cores of rotational drivers can be composed of lines of conduction block up to ~1.5 mm, suggesting microreentrant circuits as a sustaining mechanism.
  • Their investigation into spatial resolution shows that high-density catheters are necessary, and basket catheters may lead to false rotor detections.
  • The study on mean AF voltage suggests it correlates well with LGE-CMRI scar maps, indicating its potential as a functional marker.

Limitations and Future Work

  • Resolution and Coverage: A major limitation, especially in clinical settings, is the spatial and temporal resolution of current mapping catheters, which can hinder accurate driver localization.
  • Specificity of CFAEs: The paper notes that CFAEs may not be specific markers and can be generated by non-AF contributing mechanisms, requiring better analytical tools for true identification.
  • Clinical Translation: Despite advancements in analytical techniques, the success rates of therapeutic ablative approaches guided by these methods remain modest, indicating a need for further development.
  • Urgent Need for Novel Tools: The authors explicitly state the urgent need for new analytical tools to process myocardial fibrillation data more effectively.
  • Future Work: The paper implicitly points towards the need for higher-resolution mapping technologies, more robust algorithms for identifying specific mechanisms (like microreentry), and better integration of substrate and driver mapping for improved therapeutic outcomes. The novel conduction block algorithm is presented as a step in this direction.

Summary updated

BibTeX
@article{handa2018analytical,
  author = {Balvinder S. Handa and Caroline H. Roney and Charles Houston and Norman A. Qureshi and Xinyang Li and David S. Pitcher and Rasheda A. Chowdhury and Phang Boon Lim and Emmanuel Dupont and Steven A. Niederer and Chris D. Cantwell and Nicholas S. Peters and Fu Siong Ng},
  title = {{Analytical approaches for myocardial fibrillation signals}},
  journal = {Computers in Biology and Medicine},
  year = {2018},
  doi = {10.1016/j.compbiomed.2018.07.008},
}
Figure