Baseline wander removal from ECG signal using State Space Recursive Least Squares (SSRLS) adaptive filter

Removal of Baseline Wander (BLW) is the primary step in ECG signal processing. Adaptive noise cancellation (ANC) has been a popular technique for removal of noise from ECG. Conventional ANC techniques require reference signal for extraction of clean ECG. This paper investigates removal of BLW from ECG signals using State Space Recursive Least Squares (SSRLS) adaptive filter. Tracking of BLW by SSRLS has been done by using sinusoidal model in state space. Intelligent DFT has been applied for precise tuning of SSRLS filter by estimating the frequency BLW in range of 0.15 ~ 0.30 Hz. The overall scheme does not require reference signal for ECG de-noising. Successful removal of BLW from noisy ECG has been accomplished.