Discrete Wavelet Transform (DWT)

Hi,

I am working on ECG signal processing to detect atrial fibrillation and I am still quite new at the area. I have downloaded data from the MIT-BIH Arrhythmia Database where the ECG signals are sampled at 360 Hz.

For the preprocessing, I want to remove baseline wander (using DWT) which I have defined as frequencies up to 0.5 Hz. Is it correctly understood that I should decompose the signal into several levels (say 10 levels), remove the frequencies below 0.5 and then reconstruct the signal?

I was thinking about this: Decompose the signal into 10 levels,

Level 1: [180; 360[

Level 2: [90; 180[

Level 3: [45; 90[

Level 4: [22.5; 45[

Level 5: [11.25; 22.5[

Level 6: [5.625; 11.25[

Level 7: [2.8125; 5.625[

Level 8: [1.40625; 2.8125[

Level 9: [0.703125; 1.40625[

Level 10: [0.35...; 0.70...[

And remove the frequencies from level 10. Then, I should be able to reconstruct the signal (inverse DWT) using all the remaining levels...

Is it correctly understood or am I doing something wrong?

Answer 1

[Dinesh Yadav]

Yes you have understood it correctly, go ahead.