Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
126 tokens/sec
GPT-4o
47 tokens/sec
Gemini 2.5 Pro Pro
43 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
47 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Validation and Generalizability of Self-Supervised Image Reconstruction Methods for Undersampled MRI (2201.12535v3)

Published 29 Jan 2022 in eess.IV, cs.AI, and physics.med-ph

Abstract: Deep learning methods have become the state of the art for undersampled MR reconstruction. Particularly for cases where it is infeasible or impossible for ground truth, fully sampled data to be acquired, self-supervised machine learning methods for reconstruction are becoming increasingly used. However potential issues in the validation of such methods, as well as their generalizability, remain underexplored. In this paper, we investigate important 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. Two self-supervised algorithms based on self-supervised denoising and the deep image prior 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. We show that prospective reconstructions can exhibit significant distortion relative to retrospective reconstructions/ground truth. Furthermore, pixel-wise quantitative metrics may not capture differences in perceptual quality accurately, in contrast to a perceptual metric. In addition, all methods showed potential for generalization; however, generalizability is more affected by changes in anatomy/contrast than other changes. We further showed that no-reference image metrics correspond well with human rating of image quality for studying generalizability. Finally, we showed that a well-tuned compressed sensing reconstruction and learned denoising perform similarly on all data.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (48)
  1. Ssfd: Self-supervised feature distance as an mr image reconstruction quality metric. NeurIPS 2021 Workshop on Deep Learning and Inverse Problems, 2021.
  2. Unsupervised deep learning methods for biological image reconstruction. arXiv preprint arXiv:2105.08040, 2021.
  3. On instabilities of deep learning in image reconstruction and the potential costs of ai. Proceedings of the National Academy of Sciences, 117(48):30088–30095, 2020.
  4. Noise2self: Blind denoising by self-supervision. Proceedings of the International Conference on Machine Learning, pages 524–533, 2019.
  5. The perception-distortion tradeoff. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 6228–6237, 2018.
  6. Robert Bridson. Fast poisson disk sampling in arbitrary dimensions. SIGGRAPH sketches, 10(1):1, 2007.
  7. Accelerated mri with un-trained neural networks. IEEE Transactions on Computational Imaging, 7:724–733, 2021.
  8. Measuring robustness in deep learning based compressive sensing. International Conference on Machine Learning, 139:2433–2444, 2021. URL http://proceedings.mlr.press/v139/darestani21a.html.
  9. Skm-tea: A dataset for accelerated mri reconstruction with dense image labels for quantitative clinical evaluation. Thirty-fifth Conference on Neural Information Processing Systems Datasets and Benchmarks Track (Round 2), 2021.
  10. Creation of fully sampled mr data repository for compressed sensing of the knee. Proceedings of Society for MR Radiographers and Technologists, 22nd Annual Meeting. Salt Lake City, Utah, USA., 2013.
  11. Generalized autocalibrating partially parallel acquisitions (grappa). Magnetic Resonance in Medicine, 47(6):1202–1210, 2002.
  12. Chapter 2 - machine learning for image reconstruction. In S. Kevin Zhou, Daniel Rueckert, and Gabor Fichtinger, editors, Handbook of Medical Image Computing and Computer Assisted Intervention, The Elsevier and MICCAI Society Book Series, pages 25–64. Academic Press, 2020. ISBN 978-0-12-816176-0. doi: https://doi.org/10.1016/B978-0-12-816176-0.00007-7. URL https://www.sciencedirect.com/science/article/pii/B9780128161760000077.
  13. Learning a variational network for reconstruction of accelerated mri data. Magnetic Resonance in Medicine, 79(6):3055–3071, 2018.
  14. Systematic evaluation of iterative deep neural networks for fast parallel mri reconstruction with sensitivity-weighted coil combination. Magnetic Resonance in Medicine, 86(4):1859–1872, 2021. doi: https://doi.org/10.1002/mrm.28827. URL https://onlinelibrary.wiley.com/doi/abs/10.1002/mrm.28827.
  15. Deep decoder: Concise image representations from untrained non-convolutional networks. Proceedings of the International Conference on Learning Representations, 2019. URL https://openreview.net/forum?id=rylV-2C9KQ.
  16. Compressive sensing with un-trained neural networks: Gradient descent finds a smooth approximation. Proceedings of the International Conference on Machine Learning, pages 4149–4158, 2020.
  17. Perceptual losses for real-time style transfer and super-resolution. Proceedings of the European conference on computer vision, pages 694–711, 2016.
  18. Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980, 2014.
  19. Assessment of the generalization of learned image reconstruction and the potential for transfer learning. Magnetic Resonance in Medicine, 81(1):116–128, 2019.
  20. Deep-learning methods for parallel magnetic resonance imaging reconstruction: A survey of the current approaches, trends, and issues. IEEE Signal Processing Magazine, 37(1):128–140, 2020a.
  21. fastmri: A publicly available raw k-space and dicom dataset of knee images for accelerated mr image reconstruction using machine learning. Radiology: Artificial Intelligence, 2(1):e190007, 2020b.
  22. Unsupervised mri reconstruction via zero-shot learned adversarial transformers. IEEE Transactions on Medical Imaging, 2022.
  23. Rare: Image reconstruction using deep priors learned without groundtruth. IEEE Journal of Selected Topics in Signal Processing, 14(6):1088–1099, 2020.
  24. NTIRE 2020 challenge on real-world image super-resolution: Methods and results. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR Workshops, pages 2058–2076, 2020. doi: 10.1109/CVPRW50498.2020.00255. URL https://openaccess.thecvf.com/content_CVPRW_2020/html/w31/Lugmayr_NTIRE_2020_Challenge_on_Real-World_Image_Super-Resolution_Methods_and_Results_CVPRW_2020_paper.html.
  25. Sparse mri: The application of compressed sensing for rapid mr imaging. Magnetic Resonance in Medicine, 58(6):1182–1195, 2007.
  26. Blind/referenceless image spatial quality evaluator. Proceedings of the Forty Fifth ASILOMAR Conference on Signals, Systems and Computers, pages 723–727, 2011.
  27. Results of the 2020 fastmri challenge for machine learning mr image reconstruction. IEEE Transactions on Medical Imaging, 40(9):2306–2317, 2021.
  28. Accelerated mp2rage imaging using cartesian phyllotaxis readout and compressed sensing reconstruction. Magnetic Resonance in Medicine, 84(4):1881–1894, 2020.
  29. Sigpy: a python package for high performance iterative reconstruction. Proceedings of the International Society of Magnetic Resonance in Medicine, Montréal, QC, 4819, 2019.
  30. Pytorch: An imperative style, high-performance deep learning library. Advances in neural information processing systems, 32:8026–8037, 2019.
  31. Sense: sensitivity encoding for fast mri. Magnetic Resonance in Medicine, 42(5):952–962, 1999.
  32. Advances in sensitivity encoding with arbitrary k-space trajectories. Magnetic Resonance in Medicine, 46(4):638–651, 2001.
  33. Using deep learning to accelerate knee mri at 3 t: results of an interchangeability study. American journal of Roentgenology, 215(6):1421, 2020.
  34. U-net: Convolutional networks for biomedical image segmentation. In International Conference on Medical image computing and computer-assisted intervention, pages 234–241. Springer, 2015.
  35. Mri t2 mapping of the knee providing synthetic morphologic images: comparison to conventional turbo spin-echo mri. Radiology, 293(3):620–630, 2019.
  36. David Salomon. Data compression: the complete reference. Springer Science & Business Media, 2004.
  37. Deep admm-net for compressive sensing mri. Advances in neural information processing systems, 29, 2016.
  38. Espirit—an eigenvalue approach to autocalibrating parallel mri: where sense meets grappa. Magnetic Resonance in Medicine, 71(3):990–1001, 2014.
  39. Deep image prior. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 9446–9454, 2018.
  40. Blind image quality evaluation using perception based features. Proceedings of the Twenty First National Conference on Communications (NCC), pages 1–6, 2015.
  41. High-fidelity reconstruction with instance-wise discriminative feature matching loss. Proc. Intl. Soc. Mag. Reson. Med. 28, 2019.
  42. No-reference perceptual quality assessment of jpeg compressed images. Proceedings of the International Conference on Image Processing, 1:I–I, 2002.
  43. Image quality assessment: from error visibility to structural similarity. IEEE Transactions on Image Processing, 13(4):600–612, 2004.
  44. No-reference image quality metrics for structural mri. Neuroinformatics, 4(3):243–262, 2006.
  45. Self-supervised learning of physics-guided reconstruction neural networks without fully sampled reference data. Magnetic Resonance in Medicine, 84(6):3172–3191, 2020.
  46. Instabilities in conventional multi-coil mri reconstruction with small adversarial perturbations. In 2021 55th Asilomar Conference on Signals, Systems, and Computers, pages 895–899. IEEE.
  47. Can signal-to-noise ratio perform as a baseline indicator for medical image quality assessment. IEEE Access, 6:11534–11543, 2018.
  48. fastmri+: Clinical pathology annotations for knee and brain fully sampled multi-coil mri data. arXiv preprint arXiv:2109.03812, 2021.
Citations (5)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now