Code excited linear prediction codec for electrocardiogram

Abstract

In this paper we propose a CELP ECG codec for medical telemetry. The encoding algorithm is based on CODE-EXCITED LINEAR PREDICTION (CELP). The general framework proposed is: QRS detection, calculation of LPC parameter, generation of residual error signal, codebook generation, MSE (mean square error) search. The codebook is generated for residual error. The indices of the codebook and corresponding LPC parameters are transmitted where the minimum MSE occurs. A replica of the transmitter codebook is present at the receiver. Corresponding to the received index value residual error coefficients are retrieved from the receiver codebook. The ECG signal is reconstructed from the retrieved code word.

Automated Screening of Arrhythmia Using Wavelet Based Machine Learning Techniques

“…There are many methods available in the literature for registration of ECG, Pan and Tompkins [5] work is simplest, even the wavelet based technique using quadratic spline wavelet [3] is computationally exhaustive. Here in our analysis Pan Tompkins algorithm is been used for detecting R point for ECG registration [14,25]. It uses derivative, smoothing, summing and threshold operators in its architecture.…”