{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,27]],"date-time":"2026-06-27T10:33:59Z","timestamp":1782556439318,"version":"3.54.5"},"reference-count":49,"publisher":"Wiley","issue":"6","license":[{"start":{"date-parts":[[2024,2,28]],"date-time":"2024-02-28T00:00:00Z","timestamp":1709078400000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by-nc-nd\/4.0\/"}],"funder":[{"DOI":"10.13039\/100000002","name":"National Institutes of Health","doi-asserted-by":"publisher","award":["R01 HL142042"],"award-info":[{"award-number":["R01 HL142042"]}],"id":[{"id":"10.13039\/100000002","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100000002","name":"National Institutes of Health","doi-asserted-by":"publisher","award":["R01 DE12101"],"award-info":[{"award-number":["R01 DE12101"]}],"id":[{"id":"10.13039\/100000002","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100000002","name":"National Institutes of Health","doi-asserted-by":"publisher","award":["UL1\u2010RR024979"],"award-info":[{"award-number":["UL1\u2010RR024979"]}],"id":[{"id":"10.13039\/100000002","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["aapm.onlinelibrary.wiley.com"],"crossmark-restriction":true},"short-container-title":["Medical Physics"],"published-print":{"date-parts":[[2024,6]]},"abstract":"<jats:title>Abstract<\/jats:title><jats:sec><jats:title>Background<\/jats:title><jats:p>Osteoporosis is a bone disease related to increased bone loss and fracture\u2010risk. The variability in bone strength is partially explained by bone mineral density (BMD), and the remainder is contributed by bone microstructure. Recently, clinical CT has emerged as a viable option for in vivo bone microstructural imaging. Wide variations in spatial\u2010resolution and other imaging features among different CT scanners add inconsistency to derived bone microstructural metrics, urging the need for harmonization of image data from different scanners.<\/jats:p><\/jats:sec><jats:sec><jats:title>Purpose<\/jats:title><jats:p>This paper presents a new deep learning (DL) method for the harmonization of bone microstructural images derived from low\u2010 and high\u2010resolution CT scanners and evaluates the method's performance at the levels of image data as well as derived microstructural metrics.<\/jats:p><\/jats:sec><jats:sec><jats:title>Methods<\/jats:title><jats:p>We generalized a three\u2010dimensional (3D) version of GAN\u2010CIRCLE that applies two generative adversarial networks (GANs) constrained by the identical, residual, and cycle learning ensemble (CIRCLE). Two GAN modules simultaneously learn to map low\u2010resolution CT (LRCT) to high\u2010resolution CT (HRCT) and vice versa. Twenty volunteers were recruited. LRCT and HRCT scans of the distal tibia of their left legs were acquired. Five\u2010hundred pairs of LRCT and HRCT image blocks of  voxels were sampled for each of the twelve volunteers and used for training in supervised as well as unsupervised setups. LRCT and HRCT images of the remaining eight volunteers were used for evaluation. LRCT blocks were sampled at 32 voxel intervals in each coordinate direction and predicted HRCT blocks were stitched to generate a predicted HRCT image.<\/jats:p><\/jats:sec><jats:sec><jats:title>Results<\/jats:title><jats:p>Mean\u00a0\u00b1\u00a0standard deviation of structural similarity (SSIM) values between predicted and true HRCT using both 3DGAN\u2010CIRCLE\u2010based supervised (0.84\u00a0\u00b1\u00a00.03) and unsupervised (0.83\u00a0\u00b1\u00a00.04) methods were significantly (<jats:italic>p<\/jats:italic>\u00a0&lt;\u00a00.001) higher than the mean SSIM value between LRCT and true HRCT (0.75\u00a0\u00b1\u00a00.03). All Tb measures derived from predicted HRCT by the supervised 3DGAN\u2010CIRCLE showed higher agreement (CCC\u00a0 [0.956 0.991]) with the reference values from true HRCT as compared to LRCT\u2010derived values (CCC\u00a0 [0.732 0.989]). For all Tb measures, except Tb plate\u2010width (CCC\u00a0=\u00a00.866), the unsupervised 3DGAN\u2010CIRCLE showed high agreement (CCC\u00a0 [0.920 0.964]) with the true HRCT\u2010derived reference measures. Moreover, Bland\u2010Altman plots showed that supervised 3DGAN\u2010CIRCLE predicted HRCT reduces bias and variability in residual values of different Tb measures as compared to LRCT and unsupervised 3DGAN\u2010CIRCLE predicted HRCT. The supervised 3DGAN\u2010CIRCLE method produced significantly improved performance (<jats:italic>p<\/jats:italic>\u00a0&lt;\u00a00.001) for all Tb measures as compared to the two DL\u2010based supervised methods available in the literature.<\/jats:p><\/jats:sec><jats:sec><jats:title>Conclusions<\/jats:title><jats:p>3DGAN\u2010CIRCLE, trained in either unsupervised or supervised fashion, generates HRCT images with high structural similarity to the reference true HRCT images. The supervised 3DGAN\u2010CIRCLE improves agreements of computed Tb microstructural measures with their reference values and outperforms the unsupervised 3DGAN\u2010CIRCLE. 3DGAN\u2010CIRCLE offers a viable DL solution to retrospectively improve image resolution, which may aid in data harmonization in multi\u2010site longitudinal studies where scanner mismatch is unavoidable.<\/jats:p><\/jats:sec>","DOI":"10.1002\/mp.17003","type":"journal-article","created":{"date-parts":[[2024,2,28]],"date-time":"2024-02-28T07:24:15Z","timestamp":1709105055000},"page":"4258-4270","update-policy":"https:\/\/doi.org\/10.1002\/crossmark_policy","source":"Crossref","is-referenced-by-count":6,"title":["Deep learning\u2010based harmonization of trabecular bone microstructures between high\u2010 and low\u2010resolution CT imaging"],"prefix":"10.1002","volume":"51","author":[{"given":"Indranil","family":"Guha","sequence":"first","affiliation":[{"name":"Department of Electrical and Computer Engineering College of Engineering University of Iowa  Iowa City Iowa USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Syed Ahmed","family":"Nadeem","sequence":"additional","affiliation":[{"name":"Department of Radiology Carver College of Medicine University of Iowa  Iowa City Iowa USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xiaoliu","family":"Zhang","sequence":"additional","affiliation":[{"name":"Department of Electrical and Computer Engineering College of Engineering University of Iowa  Iowa City Iowa USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Paul A.","family":"DiCamillo","sequence":"additional","affiliation":[{"name":"Department of Radiology Carver College of Medicine University of Iowa  Iowa City Iowa USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Steven M.","family":"Levy","sequence":"additional","affiliation":[{"name":"Department of Preventive and Community Dentistry University of Iowa  Iowa City Iowa USA"},{"name":"Department of Epidemiology University of Iowa  Iowa City Iowa USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ge","family":"Wang","sequence":"additional","affiliation":[{"name":"Biomedical Imaging Center BME\/CBIS Rensselaer Polytechnic Institute  Troy New York USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Punam K.","family":"Saha","sequence":"additional","affiliation":[{"name":"Department of Electrical and Computer Engineering College of Engineering University of Iowa  Iowa City Iowa USA"},{"name":"Department of Radiology Carver College of Medicine University of Iowa  Iowa City Iowa USA"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"311","published-online":{"date-parts":[[2024,2,28]]},"reference":[{"key":"e_1_2_7_2_1","unstructured":"WHO Scientific Group. Prevention and management of Osteoporosis World Health Organization.2003."},{"key":"e_1_2_7_3_1","doi-asserted-by":"publisher","DOI":"10.1016\/0895-4356(88)90036-4"},{"key":"e_1_2_7_4_1","doi-asserted-by":"publisher","DOI":"10.1097\/00002142-200210000-00005"},{"key":"e_1_2_7_5_1","doi-asserted-by":"publisher","DOI":"10.1007\/BF02554913"},{"key":"e_1_2_7_6_1","doi-asserted-by":"publisher","DOI":"10.1172\/JCI111096"},{"key":"e_1_2_7_7_1","doi-asserted-by":"publisher","DOI":"10.1359\/jbmr.2000.15.1.13"},{"key":"e_1_2_7_8_1","doi-asserted-by":"publisher","DOI":"10.1007\/BF01623011"},{"key":"e_1_2_7_9_1","doi-asserted-by":"publisher","DOI":"10.1359\/jbmr.1998.13.7.1175"},{"key":"e_1_2_7_10_1","first-page":"1089","article-title":"Adaptations in trabecular bone microarchitecture in Olympic athletes determined by 7T MRI","volume":"27","author":"Chang G","year":"2008","journal-title":"J Magn Reson Med"},{"key":"e_1_2_7_11_1","doi-asserted-by":"publisher","DOI":"10.1210\/jc.2005-1258"},{"key":"e_1_2_7_12_1","doi-asserted-by":"publisher","DOI":"10.1007\/s00198-009-0913-2"},{"key":"e_1_2_7_13_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.rcl.2010.02.015"},{"key":"e_1_2_7_14_1","doi-asserted-by":"publisher","DOI":"10.1002\/mp.12632"},{"key":"e_1_2_7_15_1","doi-asserted-by":"publisher","DOI":"10.1002\/mp.15629"},{"key":"e_1_2_7_16_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.jocd.2007.12.013"},{"key":"e_1_2_7_17_1","doi-asserted-by":"publisher","DOI":"10.1109\/TMI.2013.2282370"},{"key":"e_1_2_7_18_1","doi-asserted-by":"publisher","DOI":"10.1118\/1.2789499"},{"key":"e_1_2_7_19_1","doi-asserted-by":"publisher","DOI":"10.1109\/83.491321"},{"key":"e_1_2_7_20_1","doi-asserted-by":"publisher","DOI":"10.1109\/TIP.2011.2108306"},{"key":"e_1_2_7_21_1","doi-asserted-by":"publisher","DOI":"10.1038\/s41598-018-27261-z"},{"key":"e_1_2_7_22_1","doi-asserted-by":"publisher","DOI":"10.1109\/TIP.2015.2462113"},{"key":"e_1_2_7_23_1","doi-asserted-by":"publisher","DOI":"10.1088\/1361-6560\/aacdd4"},{"key":"e_1_2_7_24_1","doi-asserted-by":"crossref","unstructured":"YuH LiuD ShiH et\u00a0al.Computed tomography super\u2010resolution using convolutional neural networks.Proc. of 2017 IEEE International Conference on Image Processing;2017:3944\u20103948.","DOI":"10.1109\/ICIP.2017.8297022"},{"key":"e_1_2_7_25_1","doi-asserted-by":"crossref","unstructured":"ChenY ShiF ChristodoulouAG XieY ZhouZ LiD.Efficient and accurate MRI super\u2010resolution using a generative adversarial network and 3D multi\u2010level densely connected network.Proc. of International Conference on Medical Image Computing and Computer\u2010Assisted Intervention;2018:91\u201099.","DOI":"10.1007\/978-3-030-00928-1_11"},{"key":"e_1_2_7_26_1","doi-asserted-by":"crossref","unstructured":"LedigC TheisL Husz\u00e1rF et\u00a0al.Photo\u2010realistic single image super\u2010resolution using a generative adversarial network.Proc. of IEEE Conference on Computer Vision and Pattern Recognition;2017:4681\u20104690.","DOI":"10.1109\/CVPR.2017.19"},{"key":"e_1_2_7_27_1","unstructured":"WangX YuK WuS et\u00a0al.ESRGAN: enhanced super\u2010resolution generative adversarial networks.Proc. of European Conference on Computer Vision;2018:1\u201016."},{"key":"e_1_2_7_28_1","doi-asserted-by":"publisher","DOI":"10.1109\/TMI.2017.2708987"},{"key":"e_1_2_7_29_1","doi-asserted-by":"publisher","DOI":"10.1109\/TMI.2019.2922960"},{"key":"e_1_2_7_30_1","doi-asserted-by":"crossref","unstructured":"ZhuJ\u2010Y ParkT IsolaP EfrosAA.Unpaired image\u2010to\u2010image translation using cycle\u2010consistent adversarial networks.Proc. of IEEE International Conference on Computer Vision.2017:2223\u20102232.","DOI":"10.1109\/ICCV.2017.244"},{"key":"e_1_2_7_31_1","doi-asserted-by":"crossref","unstructured":"GuhaI NadeemSA YouC et\u00a0al.Deep learning based high\u2010resolution reconstruction of trabecular bone microstructures from low\u2010resolution CT scans using GAN\u2010CIRCLE.Proc. of Medical Imaging 2020: Biomedical Applications in Molecular Structural and Functional Imaging.2020;11317:113170U.","DOI":"10.1117\/12.2549318"},{"key":"e_1_2_7_32_1","doi-asserted-by":"crossref","unstructured":"GuhaI NadeemSA ZhangX LevySM TornerJC SahaPK.Unsupervised GAN\u2010CIRCLE for high\u2010resolution reconstruction of bone microstructure from low\u2010resolution CT scans.Proc. of Medical Imaging 2021: Biomedical Applications in Molecular Structural and Functional Imaging.2021;11600:331\u2010341.","DOI":"10.1117\/12.2581068"},{"key":"e_1_2_7_33_1","doi-asserted-by":"crossref","unstructured":"MeijeringEH NiessenWJ PluimJP ViergeverMA.Quantitative comparison of sinc\u2010approximating kernels for medical image interpolation.Proc. of International Conference on Medical Image Computing and Computer\u2010Assisted Intervention.1999;1679:210\u2010217.","DOI":"10.1007\/10704282_23"},{"key":"e_1_2_7_34_1","unstructured":"GoodfellowI Pouget\u2010AbadieJ MirzaM et\u00a0al.Generative adversarial nets.Proc. of Advances in Neural Information Processing Systems.2014:2672\u20102680."},{"key":"e_1_2_7_35_1","unstructured":"UlyanovD VedaldiA LempitskyV.Instance normalization: the missing ingredient for fast stylization.arXiv preprint arXiv:1607.08022 2016."},{"key":"e_1_2_7_36_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.neuroimage.2006.01.015"},{"key":"e_1_2_7_37_1","doi-asserted-by":"crossref","unstructured":"HeK ZhangX RenS SunJ.Delving deep into rectifiers: surpassing human\u2010level performance on imagenet classification.Proc. of IEEE International Conference on Computer Vision;2015:1026\u20101034.","DOI":"10.1109\/ICCV.2015.123"},{"key":"e_1_2_7_38_1","unstructured":"KingmaDP BaJ.Adam: a method for stochastic optimization.arXiv preprint arXiv:1412.6980 2014."},{"key":"e_1_2_7_39_1","doi-asserted-by":"publisher","DOI":"10.1109\/TIP.2003.819861"},{"key":"e_1_2_7_40_1","doi-asserted-by":"publisher","DOI":"10.1007\/s11760-020-01790-5"},{"key":"e_1_2_7_41_1","doi-asserted-by":"publisher","DOI":"10.1002\/mrm.27178"},{"key":"e_1_2_7_42_1","doi-asserted-by":"publisher","DOI":"10.1109\/TMI.2010.2050779"},{"key":"e_1_2_7_43_1","doi-asserted-by":"publisher","DOI":"10.1109\/TBME.2014.2313564"},{"key":"e_1_2_7_44_1","doi-asserted-by":"publisher","DOI":"10.1109\/TMI.2011.2158349"},{"key":"e_1_2_7_45_1","doi-asserted-by":"publisher","DOI":"10.1006\/pmed.2002.1113"},{"key":"e_1_2_7_46_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.amepre.2005.06.001"},{"key":"e_1_2_7_47_1","doi-asserted-by":"publisher","DOI":"10.1088\/2057-1976\/acbb0a"},{"key":"e_1_2_7_48_1","doi-asserted-by":"publisher","DOI":"10.1109\/TKDE.2009.191"},{"key":"e_1_2_7_49_1","doi-asserted-by":"publisher","DOI":"10.1109\/TMI.2014.2366792"},{"key":"e_1_2_7_50_1","doi-asserted-by":"publisher","DOI":"10.1109\/JPROC.2020.3004555"}],"container-title":["Medical Physics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/aapm.onlinelibrary.wiley.com\/doi\/pdf\/10.1002\/mp.17003","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,8,23]],"date-time":"2025-08-23T00:23:18Z","timestamp":1755908598000},"score":1,"resource":{"primary":{"URL":"https:\/\/aapm.onlinelibrary.wiley.com\/doi\/10.1002\/mp.17003"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,2,28]]},"references-count":49,"journal-issue":{"issue":"6","published-print":{"date-parts":[[2024,6]]}},"alternative-id":["10.1002\/mp.17003"],"URL":"https:\/\/doi.org\/10.1002\/mp.17003","archive":["Portico"],"relation":{},"ISSN":["0094-2405","2473-4209"],"issn-type":[{"value":"0094-2405","type":"print"},{"value":"2473-4209","type":"electronic"}],"subject":[],"published":{"date-parts":[[2024,2,28]]},"assertion":[{"value":"2023-03-21","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2024-02-12","order":2,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2024-02-28","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}