Figure 1

Simplified outline of image reconstruction and analysis pipeline. The data is first subjected to noise pre-whitening and oversampling in the readout direction is removed. After pre-whitening, retrospective gating is used to bin the data into cardiac phases. Data are then Fourier transformed to image space and the multi-channel data is combined with unmixing coefficients that are calculated from GRAPPA calibration data. The reconstruction can be repeated multiple times with added random noise using the optional pseudo replica module. After image reconstruction, ROIs are defined on the images and based on the ROI geometry, the sampling pattern, and unmixing coefficients the standard deviation of the ROI phase sum is calculated according to equation (5).