I want to do multichannel ECG data compression using multiscale PCA . Do the transformed coefficients are eigenspace.
Hi Sushant,
The following links may come handy. Please check.
Article Multi-channel ECG data compression using compressed sensing ...
https://pdfs.semanticscholar.org/6ff0/2dfcb4c9e7dc4219e00a6806eafc71fe5681.pdf
Chapter Application of Compressed Sensing (CS) for ECG Signal Compre...
Hope these help.
Thanks.
A very! clear! question... (Do you -really- think that I will read throug few links you provide..!?! (Either you are too optimistic - or think that people have no life...).
...if you just want to compress signal - google zip 7z or what alg you fancy... I suppose you did not mean that (who in his healthy mind would analyze lossy compressed signal? Or was that what you wanted... (please uncompress the data before you do anything..))
The qusestion is that what base frequency is common for all of your observation that you can remove - or is there a common pattern to frequecny that you can only submit from partial observations (though there has to be common pattern there usually or you cannot use it...)
So (I did not look at the your "check" - but) first find the correlation between the "shared" signals... Find the common behaviour with the signals and remove that as base signal (or shift them to the "common" origin and then do the subtraction - ie find the common offset to frequencies and normalize) - now you have compressible signal around the baseband and you can do magics (if you do not get it read more about signal processing) ( you might need some phase locking mechanism to find the correlating signal - I have only worked with machines with quite well defined eigen frequencies...-but heart is usually such...?)...
Really - one has to study a bit... (there is no magics to convert knowledge from one to other...
)
- Badges
- Science topic
- Similar topics
- Biological Science
- Physiology
- Biosignals
- Biomedical Signal Processing