ECG baseline drift removal using discrete wavelet transform

This study presents a method for baseline drift removal in the ECG signal by means of the discrete wavelet transform with the Daubechies-4, in where the central pseudo-frequency parameter is used. The sampling frequency of ECG record related to each decomposition scale of the transform is assessed. DWT is applied to the level of decomposition calculated with the above values; this makes the algorithm computationally efficient so it can be used at online processes. Nine ECG records of the MIT-BIH Arrhythmia database are used for method evaluation. To prove its efficiency, these records are contaminated and then processed, obtaining a correlation coefficient average of 0.99.