I am extracting the feature values(kurtosis, power, skewness, and entropy) for Alzheimer's and Healthy individuals. The values I'm getting in both cases seem to be similar. I don't know why that is the case although both the data are clearly different. The power spectrum is different but the feature values are similar.
For example,
Alzheimer's disease for T3 scalp electrode (Kurtosis= 4.33616, Entropy= 3.01776, Skewness= -0.0004604).
Healthy Individual for T3 scalp electrode( Kurtosis= 4.579, Entropy= 2.14331, Skewness= 0.0017).
I used Matlab to find out the features using the inbuilt functions.
It would be helpful for me if there was a way to differentiate the features in Matlab.
P.s If there was a paper that explains the values also would be much appreciated.
I am not an expert, but I know, that signal processing is far from easy.
I will ad a file, that gives you something to "process".
You can try to read our paper on the prediction of arrhythmias from ECG recordings. It is another biosignal generated by our bodies. We combined the Entropy measure and AI/ML machine learning methods to get the results. All is described in detail. The fully finished paper will come next year.
https://www.researchgate.net/publication/340460356_Application_of_Machine_Learning_and_Complexity_in_Medicine_Prediction_of_Drug_Induced_Arrhythmias_in_Rabbit_Model_up_to_One_Hour_Before_Their_Onset_early_version_requested_due_to_COVID-19_unfinished_m
- Badges
- Science topic