Cardiac Arrhythmia Classification Using a combination of Quadratic Spline-Based Wavelet Transform and Artificial Neural Classification Network

The authors present the use of Wavelet Transform, using a quadratic spline function, and Probabilistic Neural Network (PNN) to classify 8 heartbeat conditions. The process consists of four mains stages. The first part consists of preprocessing a nd f iltering selected ECG l ead II (D II) data re gisters f rom t he PhysioNet repository. The filtered signal is fed to a w avelet transform process using a quadratic s pline f unction, to obtain a f eature v ector. T he r esults ar e transferred to a Probabilistic Neural Network algorithm for heartbeat classification. Finally, the algorithm is tested with confusion matrices to determine classification accuracy. The algorithm yielded a 9 1.5%, 90.3% and 95.5% classification accu racy f or au ricular f ibrillation, s inoauricular h eart block an d p aroxysmal atrial fibrillation conditions respectively. The lower scores were obtained for p remature at rial co ntraction and premature v entricular co ntraction conditions (75.5% and 69.9% respectively). However, considering the validation test conditions, the results suggest the algorithm is suitable for on-line classification of heartbeat conditions as part of a DSP-based Holter device.