ECG signal reconstruction from undersampled measurement using a trained overcomplete dictionary
- Author(s): Lee, SJ
- Luan, J
- Chou, PH
- et al.
Published Web Locationhttps://doi.org/10.12988/ces.2014.411219
© 2014 Seung Jae Lee, Jun Luan, and Pai H. Chou. We propose a new approach to reconstructing ECG signal from undersampled data based on constructing a combined overcomplete dictionary. The dictionary is obtained by combining the trained dictionary by K-SVD dictionary learning algorithm with universal types of dictionary such as DCT or wavelet basis. Using the trained overcomplete dictionary, the proposed method can find sparse approximation by compressive sensing. Experimental results on MIT-BIH arrhythmia database confirm that our proposed algorithm has high reconstruction performance while maintaining low distortion.