従来の方法とディープラーニングのギャップを埋める新たな研究 New research closes the gap between traditional and deep learning methods
2022-09-14 ミネソタ大学
多くの研究により、深層学習は従来の圧縮センシングよりも大きなマージンをもって優れていることが示されています。しかし、ディープラーニングの使用にはいくつかの懸念があります。例えば、学習データが不十分だと、アルゴリズムにバイアスがかかり、MRIの結果を誤って解釈してしまう可能性があります。
最新のデータサイエンスツールと機械学習のアイデアを組み合わせることで、従来の圧縮手法を微調整し、ディープラーニングとほぼ同等の品質を実現する方法を発見した。
<関連情報>
- https://cse.umn.edu/college/news/researchers-combine-data-science-and-machine-learning-techniques-improve-traditional
- https://www.pnas.org/doi/10.1073/pnas.2201062119
Revisiting ℓ1-wavelet compressed-sensing MRI in the era of deep learning
Hongyi Gu, Burhaneddin Yaman, Steen Moeller, Jutta Ellermann, Kamil Ugurbil and Mehmet Akçakaya
Proceedings of the National Academy of Sciences Published:August 8, 2022
DOI:https://doi.org/10.1073/pnas.2201062119
Abstract
Following their success in numerous imaging and computer vision applications, deep-learning (DL) techniques have emerged as one of the most prominent strategies for accelerated MRI reconstruction. These methods have been shown to outperform conventional regularized methods based on compressed sensing (CS). However, in most comparisons, CS is implemented with two or three hand-tuned parameters, while DL methods enjoy a plethora of advanced data science tools. In this work, we revisit ℓ1-wavelet CS reconstruction using these modern tools. Using ideas such as algorithm unrolling and advanced optimization methods over large databases that DL algorithms utilize, along with conventional insights from wavelet representations and CS theory, we show that ℓ1-wavelet CS can be fine-tuned to a level close to DL reconstruction for accelerated MRI. The optimized ℓ1-wavelet CS method uses only 128 parameters compared to >500,000 for DL, employs a convex reconstruction at inference time, and performs within <1% of a DL approach that has been used in multiple studies in terms of quantitative quality metrics.
