Originally published in:
https://doi.org/10.1148/radiol.230764
How AI May Transform Musculoskeletal Imaging
Ali Guermazi, Patrick Omoumi, Mickael Tordjman, Jan Fritz, Richard Kijowski, Nor-Eddine Regnard, John Carrino, Charles E. Kahn Jr, Florian Knoll, Daniel Rueckert, Frank W. Roemer, Daichi Hayash
Erratum in:
https://doi.org/10.1148/radiol.249002
Errors appeared in the legend and caption of Figure 1. The images in panels E and J were incorrectly labeled as fourth place; these images actually show the non–deep learning reference reconstructions that were performed by the challenge organizers after the conclusion of the challenge. The images in E and J were therefore relabeled as non–deep learning reference. In the legend, the sentence “The first, second, third, and fourth place images each indicate deep learning (DL) reconstructions…” has been amended to “The first, second, and third place images each indicate deep learning (DL) reconstructions…” As well, (B–E) was changed to (B–D) and (G–J) was changed to (G–I). Finally, a sentence was changed to include a callout to E and J, as follows: “After the conclusion of the challenge, the challenge organizers performed a combined parallel imaging and compressed sensing reconstruction as a state-of-the-art reference (E, J).”