ISP-IRLNet: Joint optimization of interpretable sampler and implicit regularization learning network for accerlerated MRI

Xing Li, Yan Yang, Hairong Zheng, Zongben Xu

Abstract: Highlights•Optimizing sample pattern and reconstruction network jointly to maintain stability.•Mixture Laplacian probabilistic distribution models the prior of sample locations.•Differentiable binarization strategy keeps approximating gradients for backward.•Implicit regularization reconstruction network approximates image prior information.