Adaptive independent component analysis to analyze electrocardiograms

Seong-Bin Yim; Harold H. Szu

doi:10.1117/12.421224

26 March 2001 Adaptive independent component analysis to analyze electrocardiograms

Seong-Bin Yim, Harold H. Szu

Proceedings Volume 4391, Wavelet Applications VIII; (2001) https://doi.org/10.1117/12.421224
Event: Aerospace/Defense Sensing, Simulation, and Controls, 2001, Orlando, FL, United States

Abstract

In this work, we apply adaptive version independent component analysis (ADAPTIVE ICA) to the nonlinear measurement of electro-cardio-graphic (ECG) signals for potential detection of abnormal conditions in the heart. In principle, unsupervised ADAPTIVE ICA neural networks can demix the components of measured ECG signals. However, the nonlinear pre-amplification and post measurement processing make the linear ADAPTIVE ICA model no longer valid. This is possible because of a proposed adaptive rectification pre-processing is used to linearize the preamplifier of ECG, and then linear ADAPTIVE ICA is used in iterative manner until the outputs having their own stable Kurtosis. We call such a new approach adaptive ADAPTIVE ICA. Each component may correspond to individual heart function, either normal or abnormal. Adaptive ADAPTIVE ICA neural networks have the potential to make abnormal components more apparent, even when they are masked by normal components in the original measured signals. This is particularly important for diagnosis well in advance of the actual onset of heart attack, in which abnormalities in the original measured ECG signals may be difficult to detect. This is the first known work that applies Adaptive ADAPTIVE ICA to ECG signals beyond noise extraction, to the detection of abnormal heart function.

Citation Download Citation

Seong-Bin Yim and Harold H. Szu "Adaptive independent component analysis to analyze electrocardiograms", Proc. SPIE 4391, Wavelet Applications VIII, (26 March 2001); https://doi.org/10.1117/12.421224

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available