Fig. 2From: Fully-automated global and segmental strain analysis of DENSE cardiovascular magnetic resonance using deep learning for segmentation and phase unwrappingData augmentation for the phase-unwrapping convolutional neural networks (CNN). Figure a shows how a new phase-wrapping pattern is generated during data augmentation using an original wrapped image as input, performing phase unwrapping, scaling of the unwrapped phase, and wrapping to the range of (− π,π). Figure b demonstrates an example of serial operations to generate augmented dataBack to article page