ICA Based Flow Artifact Removal from ECG during MRI

In this paper, a method of Electrocardiographic (ECG) signal denoising during Magnetic Resonance Imaging (MRI) is proposed. In MR environment ECG signals will get corrupted by flow artifacts and hence reduces its diagnostic value. The proposed method focuses on removal of flow artifacts from multi-channel ECG signals, thereby improving its diagnostic quality. The standard or instantaneous independent component analysis (ICA) model is used with a reference signal to eliminate the flow artifacts. The performance of method is evaluated using MGH/MF multiparameter database with simulated flow artifacts. The diagnostic quality of filtered ECG is evaluated with different quality measures. In future, the performance of the method is to be studied on ECGs recorded during the MRI examination.