ECG features are extracted using different morphological and statistical techniques. I want to know the new effective feature extraction techniques.
There are many optimisation techniques recently published. However, there are evidences that there is no difference among these algorithms. I think that using the old style optimisation methods is better. Try to use PSO, GA, Simulated annealing, and the other mature algorithms.
There are many feature extraction techniques (MMFS, PCA, MPCA, Neural Networks, GA) The main differences between them lies on the computing costs (massive CPU times). I enclose for you one of the thesis that applies some of these feature extraction techniques on ECG signals.
Artificial Neural networks, Annealing, Wavelet and Fast ICA also have been used at many instances
I'ld start to read "The principle of Software QRS Detection" Bert-Uwe Köhler, Carsten Hennig, Reinhold Orglmeister.
- Badges
- Science topic
- Similar topics
- Biological Science
- Physiology
- Biosignals
- Biomedical Signal Processing