An objective tool for quantifying atrial fibrillation substrate in rats.

The utility of rodents for research related to atrial fibrillation (AF) is growing exponentially. However, the obtained arrhythmic waveforms are often mixed with ventricular signals and the ability to analyze regularity and complexity of such events is limited. Recently, we introduced an implantable quadripolar electrode adapted for advanced atrial electrophysiology in ambulatory rats. Notably, we have found that the implantation itself leads to progressive atrial remodeling, presumably because of mechanical loading of the atria. In the present study, we developed an algorithm to clean the atrial signals from ventricular mixing and thereafter quantify the AF substrate in an objective manner based on waveform complexity. Rats were sequentially examined 1-, 4-, and 8-wk postelectrode implantation using a standard AF triggering protocol. Preburst ventricular mixing was sampled and automatically subtracted based on QRS detection in the ECG. Thereafter, the "pure" atrial signals were analyzed by Lempel-Ziv complexity algorithm and a complexity ratio (CR) was defined for each signal by normalizing the postburst to the preburst values. Receiver operating characteristic (ROC) curve analysis indicated an optimal CR cutoff of 1.236 that detected irregular arrhythmic events with high sensitivity (94.5%), specificity (93.1%), and area under the curve (AUC) (0.96, 95% confidence interval, 0.945-0.976). Automated and unbiased analysis indicated a gradual increase in signal complexity over time with augmentation of high frequencies in power spectrum analysis. Our findings indicate that CR algorithm detects irregularity in a highly efficient manner and can also detect the atrial remodeling induced by electrode implantation. Thus, CR analysis can strongly facilitate standardized AF research in rodents. Rodents are increasingly used in AF research. However, because of technical difficulties including atrial waveform mixing by ventricular signals, most studies do not discriminate between irregular (i.e., AF) and regular atrial arrhythmias. Here, we develop an unbiased computerized tool to "pure" the atrial signals from ventricular mixing and thereafter analyze AF substrate based on the level of irregularity in an objective manner. This novel tool can facilitate standardized AF research in rodents.