Atrial Flutter Detection System by AdEx Encoded Lead-II ECG

Quantitatively characterizing atrial electrical activity during irregular heart rhythms, such as atrial fibrillation (AF) and atrial flutter (AFL), is challenging. In this work, a novel time-domain system for detecting AFL has been implemented using a Spiking Neuron Model on an FPGA. This approach leverages a biologically plausible AdEx neuron model to encode the data. Parameters such as Individual Flutter Width (IFW) and Atrial Cycle Length (ACL) were extracted from the encoded spikes. The system was tested with ECG signals corresponding to Modified Lead-II (ML-II) from the MIT-BIH database, achieving sensitivities of 92% for record 202 from the MIT-BIH Arrhythmia Database and 88% for record 04043 from the MIT-BIH Atrial Fibrillation Database.