Deep-learning synthesized pseudo-CT for MR high-resolution pediatric cranial bone imaging (MR-HiPCB).
Deep-learning synthesized pseudo-CT for MR high-resolution pediatric cranial bone imaging (MR-HiPCB).
复制标题
DOI:
10.1002/mrm.29356
复制
发表时间:
2022-11
影响因子:
3.3
通讯作者:
An, Hongyu
中科院分区:
文献类型:
--
作者:
Boroojeni, Parna Eshraghi;Chen, Yasheng;Commean, Paul K.;Eldeniz, Cihat;Skolnick, Gary B.;Merrill, Corinne;Patel, Kamlesh B.;An, Hongyu
CT is routinely used to detect cranial abnormalities in pediatric patients with head trauma or craniosynostosis. This study aimed to develop a deep learning method to synthesize pseudo-CT (pCT) images for MRI High-resolution Pediatric Cranial Bone imaging (MR-HiPCB) to eliminating ionizing radiation from CT. 3D Golden-Angle stack-of-stars MRI were obtained from 44 pediatric participants. Two patch-based ResUNets were trained using paired MR and CT patches randomly selected from the whole head (NetWH) or in the vicinity of bone, fractures/sutures, or air (NetBA) to synthesize pCT. A third ResUNet was trained to generate a binary brain mask using only MRI. The pCT images from NetWH (pCTNetWH) in the brain area and NetBA (pCTNetBA) in the non-brain area were combined to generate pCTCom. A manual processing method using inverted MR images (iMR) was also employed for comparison. pCTCom (68.01±14.83 HU) had significantly smaller mean absolute errors (MAE) than pCTNetWH (82.58±16.98 HU, P<0.0001) and pCTNetBA (91.32±17.2 HU, P<0.0001) in the whole head. Within cranial bone, the MAE of pCTCom (227.92±46.88 HU) was significantly lower than pCTNetWH (287.85±59.46 HU, P<0.0001) but similar to pCTNetBA (230.20±46.17 HU). Dice Similarity Coefficient (DSC) of the segmented bone was significantly higher in pCTCom (0.90±0.02) than in pCTNetWH (0.86±0.04, P<0.0001), pCTNetBA (0.88±0.03, P<0.0001), and iMR (0.71±0.09, P<0.0001). DSC from pCTCom demonstrated significantly reduced age dependence than iMR. Furthermore, pCTCom provided excellent suture and fracture visibility comparable to CT. MR-HiPCB may facilitate the clinical translation of a radiation-free MR cranial bone imaging method for pediatric patients.
登录
查看更多内容
影响因子:
--
作者:
Dorfman, Adam L.;Fazel, Reza;Einstein, Andrew J.;Applegate, Kimberly E.;Krumholz, Harlan M.;Wang, Yongfei;Christodoulou, Emmanuel;Chen, Jersey;Sanchez, Ramon;Nallamothu, Brahmajee K.
通讯作者:
Nallamothu, Brahmajee K.
影响因子:
9.3
作者:
Hofmann, Matthias;Steinke, Florian;Pichler, Bernd J.
通讯作者:
Pichler, Bernd J.
影响因子:
2.5
作者:
HAACKE E M;PATRICK J L
通讯作者:
PATRICK J L
影响因子:
19.7
作者:
Chen Y;Juttukonda M;Su Y;Benzinger T;Rubin BG;Lee YZ;Lin W;Shen D;Lalush D;An H
通讯作者:
An H
影响因子:
3.8
作者:
Johansson, Adam;Karlsson, Mikael;Nyholm, Tufve
通讯作者:
Nyholm, Tufve