Share Email Print
cover

Proceedings Paper

Towards non-invasive electrocardiographic imaging using regularized neural networks
Author(s): Abhejit Rajagopal; Vincent Radzicki; Hua Lee; Shivkumar Chandrasekaran
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

We present a new data-driven technique for non-invasive electronic imaging of cardiovascular tissues using routinely-measured body-surface electrocardiogram (ECG) signals. While traditional ECG imaging and 3D reconstruction algorithms typically rely on a combination of linear Fourier theory, geometric and parametric modeling, and invasive measurements via catheters, we show in this work that it is possible to learn the complicated inverse map, from body-surface potentials to epicardial or endocardial potentials, by exploiting the powerful approximation properties of neural networks. The key contribution here is a formulation of the inverse problem that allows historical data to be leveraged as ground-truth for training the inverse operator. We provide some initial experiments, and outline a path for extending this technique for real-time diagnostic applications.

Paper Details

Date Published: 9 March 2018
PDF: 9 pages
Proc. SPIE 10573, Medical Imaging 2018: Physics of Medical Imaging, 105732O (9 March 2018); doi: 10.1117/12.2294474
Show Author Affiliations
Abhejit Rajagopal, Univ. of California, Santa Barbara (United States)
Vincent Radzicki, Univ. of California, Santa Barbara (United States)
Hua Lee, Univ. of California, Santa Barbara (United States)
Shivkumar Chandrasekaran, Univ. of California, Santa Barbara (United States)


Published in SPIE Proceedings Vol. 10573:
Medical Imaging 2018: Physics of Medical Imaging
Joseph Y. Lo; Taly Gilat Schmidt; Guang-Hong Chen, Editor(s)

© SPIE. Terms of Use
Back to Top