TY - JOUR
T1 - An untrained deep learning method for reconstructing dynamic MR images from accelerated model-based data
AU - Slavkova, Kalina P.
AU - DiCarlo, Julie C.
AU - Wadhwa, Viraj
AU - Kumar, Sidharth
AU - Wu, Chengyue
AU - Virostko, John
AU - Yankeelov, Thomas E.
AU - Tamir, Jonathan I.
N1 - Publisher Copyright:
© 2022 International Society for Magnetic Resonance in Medicine.
PY - 2023/4
Y1 - 2023/4
N2 - Purpose: To implement physics-based regularization as a stopping condition in tuning an untrained deep neural network for reconstructing MR images from accelerated data. Methods: The ConvDecoder (CD) neural network was trained with a physics-based regularization term incorporating the spoiled gradient echo equation that describes variable-flip angle data. Fully-sampled variable-flip angle k-space data were retrospectively accelerated by factors of R = {8, 12, 18, 36} and reconstructed with CD, CD with the proposed regularization (CD + r), locally low-rank (LR) reconstruction, and compressed sensing with L1-wavelet regularization (L1). Final images from CD + r training were evaluated at the “argmin” of the regularization loss; whereas the CD, LR, and L1 reconstructions were chosen optimally based on ground truth data. The performance measures used were the normalized RMS error, the concordance correlation coefficient, and the structural similarity index. Results: The CD + r reconstructions, chosen using the stopping condition, yielded structural similarity indexs that were similar to the CD (p = 0.47) and LR structural similarity indexs (p = 0.95) across R and that were significantly higher than the L1 structural similarity indexs (p = 0.04). The concordance correlation coefficient values for the CD + r T1 maps across all R and subjects were greater than those corresponding to the L1 (p = 0.15) and LR (p = 0.13) T1 maps, respectively. For R ≥ 12 (≤4.2 min scan time), L1 and LR T1 maps exhibit a loss of spatially refined details compared to CD + r. Conclusion: The use of an untrained neural network together with a physics-based regularization loss shows promise as a measure for determining the optimal stopping point in training without relying on fully-sampled ground truth data.
AB - Purpose: To implement physics-based regularization as a stopping condition in tuning an untrained deep neural network for reconstructing MR images from accelerated data. Methods: The ConvDecoder (CD) neural network was trained with a physics-based regularization term incorporating the spoiled gradient echo equation that describes variable-flip angle data. Fully-sampled variable-flip angle k-space data were retrospectively accelerated by factors of R = {8, 12, 18, 36} and reconstructed with CD, CD with the proposed regularization (CD + r), locally low-rank (LR) reconstruction, and compressed sensing with L1-wavelet regularization (L1). Final images from CD + r training were evaluated at the “argmin” of the regularization loss; whereas the CD, LR, and L1 reconstructions were chosen optimally based on ground truth data. The performance measures used were the normalized RMS error, the concordance correlation coefficient, and the structural similarity index. Results: The CD + r reconstructions, chosen using the stopping condition, yielded structural similarity indexs that were similar to the CD (p = 0.47) and LR structural similarity indexs (p = 0.95) across R and that were significantly higher than the L1 structural similarity indexs (p = 0.04). The concordance correlation coefficient values for the CD + r T1 maps across all R and subjects were greater than those corresponding to the L1 (p = 0.15) and LR (p = 0.13) T1 maps, respectively. For R ≥ 12 (≤4.2 min scan time), L1 and LR T1 maps exhibit a loss of spatially refined details compared to CD + r. Conclusion: The use of an untrained neural network together with a physics-based regularization loss shows promise as a measure for determining the optimal stopping point in training without relying on fully-sampled ground truth data.
KW - image reconstruction
KW - inverse problem
KW - physics-guided deep learning
KW - quantitative MRI
UR - http://www.scopus.com/inward/record.url?scp=85143972774&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85143972774&partnerID=8YFLogxK
U2 - 10.1002/mrm.29547
DO - 10.1002/mrm.29547
M3 - Article
C2 - 36468624
AN - SCOPUS:85143972774
SN - 0740-3194
VL - 89
SP - 1617
EP - 1633
JO - Magnetic resonance in medicine
JF - Magnetic resonance in medicine
IS - 4
ER -