Validation and Generalizability of Self-Supervised Image Reconstruction Methods for Undersampled MRI

Purpose: To investigate aspects of the validation of self-supervised algorithms for reconstruction of undersampled MR images: quantitative evaluation of prospective reconstructions, potential differences between prospective and retrospective reconstructions, suitability of commonly used quantitative metrics, and generalizability. Theory and Methods: Two self-supervised algorithms based on self-supervised denoising and neural network image priors were investigated. These methods are compared to a least squares fitting and a compressed sensing reconstruction using in-vivo and phantom data. Their generalizability was tested with prospectively under-sampled data from experimental conditions different to the training. Results: Prospective reconstructions can exhibit significant distortion relative to retrospective reconstructions/ground truth. Pixel-wise quantitative metrics may not capture differences in perceptual quality accurately, in contrast to a perceptual metric. All methods showed potential for generalization; generalizability is more affected by changes in anatomy/contrast than other changes. No-reference image metrics correspond well with human rating of image quality for studying generalizability. Compressed Sensing and learned denoising perform similarly well on all data. Conclusion: Self-supervised methods show promising results for accelerating image reconstruction in clinical routines. Nonetheless, more work is required to investigate standardized methods to validate reconstruction algorithms for future clinical use.