Real-Time Rotational Activity Detection in Atrial Fibrillation

Abstract

Rotational activations, or spiral waves, are one of the proposed mechanisms for atrial fibrillation (AF) maintenance. We present a system for assessing the presence of rotational activity from intracardiac electrograms (EGMs). Our system is able to operate in real-time with multi-electrode catheters of different topologies in contact with the atrial wall, and it is based on new local activation time (LAT) estimation and rotational activity detection methods. The EGM LAT estimation method is based on the identification of the highest sustained negative slope of unipolar signals. The method is implemented as a linear filter whose output is interpolated on a regular grid to match any catheter topology. Its operation is illustrated on selected signals and compared to the classical Hilbert-Transform-based phase analysis. After the estimation of the LAT on the regular grid, the detection of rotational activity in the atrium is done by a novel method based on the optical flow of the wavefront dynamics, and a rotation pattern match. The methods have been validated using in silico and real AF signals.

Standardizing Single-Frame Phase Singularity Identification Algorithms and Parameters in Phase Mapping During Human Atrial Fibrillation

“…Novel computer algorithms for AF driver identification – and consequently targets for ablation – have been extensively explored to study the underlying persAF mechanisms aiming to improving ablative treatment outcomes ( Narayan et al, 2014 ; Li et al, 2017b ; Rios-Munoz et al, 2018 ). Real-time implementation of rotor detection has shown great potential ( Rios-Munoz et al, 2018 ), hence the investigation of the processing time is important for the further development of real-time EP tools to guide catheter ablation of AF. Our results show the convolutional kernel method (Algorithm 3) was faster than the neighbor-indexing algorithms (algorithms 1 and 2) – in which the latter needed a larger number of loop operations for checking the monotonic increase/decrease in phase values in loops of neighbors.…”

Early Detection of Atrial Fibrillation Based on ECG Signals

“…Comparing with the existing methods [22][23][24][25][26][27][28][29] to detect AF, the proposed method does not require complex numerical tools such as RdR maps [22] or Receiver Operating Characteristic (ROC) curves [25,28]. It also does not require computationally intensive algorithms [23,27] or probability-based data adaptive techniques [24], community-based screening [25], or mobile phone apps [29]. Instead, this research work focused on developing a noninvasive method to detect AF in a home setting based on low cost data acquisition hardware and low demand for computing power.…”

Patient-Tailored In Silico 3D Simulations and Models From Electroanatomical Maps of the Left Atrium

“…We built a database of personalized atrial models in a straightforward way. This database could be used to test clinical treatments, new signal processing techniques, and validation methods as in [9,10]. Further validation is to be done, including additional tissue properties, e.g.…”