{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,19]],"date-time":"2026-09-19T00:29:19Z","timestamp":1789777759392,"version":"4.0.1"},"reference-count":40,"publisher":"Wiley","issue":"6","license":[{"start":{"date-parts":[[2025,4,9]],"date-time":"2025-04-09T00:00:00Z","timestamp":1744156800000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/onlinelibrary.wiley.com\/termsAndConditions#vor"}],"funder":[{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["62272011"],"award-info":[{"award-number":["62272011"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["61876008"],"award-info":[{"award-number":["61876008"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100004826","name":"Natural Science Foundation of Beijing Municipality","doi-asserted-by":"publisher","award":["7232337"],"award-info":[{"award-number":["7232337"]}],"id":[{"id":"10.13039\/501100004826","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":[[2025,6]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:sec>\n                    <jats:title>Background<\/jats:title>\n                    <jats:p>Unsupervised traumatic brain injury (TBI) lesion detection aims to identify and segment abnormal regions, such as cerebral edema and hemorrhages, using only healthy training data. Recent advancements in generative models have achieved success in unsupervised anomaly detection by transforming abnormal patterns into normal counterparts. However, current mask\u2010free image generators often fail to maintain semantic consistency of anatomical structures during the restoration process. This limitation negatively impacts residual\u2010based anomaly detection, particularly in cases where structural deformations occur due to the mass effect of TBI\u00a0lesions.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Purpose<\/jats:title>\n                    <jats:p>This study aims to develop a semantic\u2010consistent, unsupervised TBI lesion detection and segmentation method that minimizes false positives by preserving normal tissue consistency during the image generation process while addressing mass effect\u2010related tissue\u00a0deformations.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>We propose the semantic\u2010consistent diffusion model (SCDM) for unsupervised TBI lesion detection, focusing on the localization and segmentation of various lesion types from noncontrast CT scans of TBI patients. Leveraging the high\u2010quality image generation capabilities of unconditioned diffusion models (DM), we introduce a normal tissue retainment (NTR) regularization to ensure that normal tissues remain unaltered throughout the iterative denoising process. Furthermore, we address normal tissue compression and deformation caused by the mass effect of TBI lesions through diffeomorphic registration, reducing erroneous activations in residual images and final lesion\u00a0maps.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>Extensive experiments were conducted on three publicly available brain lesion datasets and one internal dataset. These datasets comprised 75, 51, 92, and 56 CT scans, respectively. Thirty seven CT scans without TBI lesions were used for training and validation, while the remaining scans were used for testing. The proposed method achieved average DSC of 0.56, 0.51, 0.47, and 0.52 and AUPRC of 0.57, 0.48, 0.53, and 0.50 on the BCIHM, BHSD, Seg\u2010CQ500, and internal datasets, respectively, surpassing state\u2010of\u2010the\u2010art unsupervised methods for TBI lesion detection and segmentation. An ablation study validated the effectiveness of the proposed NTR regularization and diffeomorphic registration\u2010based mass effect\u00a0simulation.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusions<\/jats:title>\n                    <jats:p>The results suggest that the proposed SCDM enables effective TBI lesion detection and segmentation across diverse TBI CT scans. It significantly reduces false positives by addressing inconsistencies in normal tissue during the iterative image restoration process and mitigating mass effect\u2010induced tissue\u00a0deformations.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.1002\/mp.17811","type":"journal-article","created":{"date-parts":[[2025,4,9]],"date-time":"2025-04-09T23:21:02Z","timestamp":1744240862000},"page":"4498-4512","update-policy":"https:\/\/doi.org\/10.1002\/crossmark_policy","source":"Crossref","is-referenced-by-count":4,"title":["Semantic\u2010consistent diffusion model for unsupervised traumatic brain injury detection and segmentation from computed tomography images"],"prefix":"10.1002","volume":"52","author":[{"given":"Diya","family":"Sun","sequence":"first","affiliation":[{"name":"Peking University People's Hospital, Institute for Artificial Intelligence, Key Laboratory of Trauma Treatment and Neural Regeneration Peking University Beijing China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yuru","family":"Pei","sequence":"additional","affiliation":[{"name":"School of Intelligence Science and Technology, Key Laboratory of Machine Perception (MOE), State Key Laboratory of General Artificial Intelligence Peking University Beijing China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Liyi","family":"Ying","sequence":"additional","affiliation":[{"name":"School of Mathematics, Statistics and Mechanic Beijing University of Technology Beijing China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Tianbing","family":"Wang","sequence":"additional","affiliation":[{"name":"Peking University People's Hospital, Institute for Artificial Intelligence, Key Laboratory of Trauma Treatment and Neural Regeneration Peking University Beijing China"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"311","published-online":{"date-parts":[[2025,4,9]]},"reference":[{"key":"e_1_2_9_2_1","doi-asserted-by":"publisher","DOI":"10.1016\/S1474-4422(18)30415-0"},{"key":"e_1_2_9_3_1","doi-asserted-by":"publisher","DOI":"10.1109\/JBHI.2021.3103850"},{"key":"e_1_2_9_4_1","doi-asserted-by":"publisher","DOI":"10.3174\/ajnr.A5742"},{"key":"e_1_2_9_5_1","doi-asserted-by":"publisher","DOI":"10.1038\/s41598-019-54491-6"},{"key":"e_1_2_9_6_1","doi-asserted-by":"crossref","unstructured":"KwonD AhnJ KimJ et\u00a0al.Siamese U\u2010Net with healthy template for accurate segmentation of intracranial hemorrhage. In:International Conference on Medical Image Computing and Computer\u2010Assisted Inter vention.Springer;2019.","DOI":"10.1007\/978-3-030-32248-9_94"},{"key":"e_1_2_9_7_1","doi-asserted-by":"publisher","DOI":"10.1161\/STROKEAHA.119.026561"},{"key":"e_1_2_9_8_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.media.2022.102489"},{"key":"e_1_2_9_9_1","doi-asserted-by":"crossref","unstructured":"WangD WangC MastersL BarnettMH.Masked multi\u2010task network for case\u2010level intracranial hemorrhage classification in brain CT volumes. In:International Conference on Medical Image Computing and Computer\u2010Assisted Intervention.Springer;2020.","DOI":"10.1007\/978-3-030-59728-3_15"},{"key":"e_1_2_9_10_1","doi-asserted-by":"crossref","unstructured":"GuoD WeiH ZhaoP et\u00a0al.Simultaneous classification and segmentation of intracranial hemorrhage using a fully convolutional neural network. In:2020 IEEE 17th International Symposium on Biomedical Imaging.IEEE;2020:118\u2010121.","DOI":"10.1109\/ISBI45749.2020.9098596"},{"key":"e_1_2_9_11_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.compmedimag.2021.101908"},{"key":"e_1_2_9_12_1","unstructured":"KuoW H\u00e4neC YuhEL MukherjeeP MalikJ.PatchFCN for intracranial hemorrhage detection.ArXiv.2018;abs\/1806.03265."},{"key":"e_1_2_9_13_1","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-87196-3_54"},{"key":"e_1_2_9_14_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.cmpb.2020.105546"},{"key":"e_1_2_9_15_1","doi-asserted-by":"crossref","unstructured":"ToikkanenM KwonD LeeM.ReSGAN: intracranial hemorrhage segmentation with residuals of synthetic brain CT scans. In:International Conference on Medical Image Computing and Computer\u2010Assisted Intervention.Springer;2021.","DOI":"10.1007\/978-3-030-87193-2_38"},{"key":"e_1_2_9_16_1","first-page":"141","article-title":"Deep learning for medical anomaly detection\u2014a survey","volume":"54","author":"Fernando T","year":"2021","journal-title":"ACM Comput Surv"},{"key":"e_1_2_9_17_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.metrad.2024.100082"},{"key":"e_1_2_9_18_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.media.2023.102846"},{"key":"e_1_2_9_19_1","unstructured":"LopezR BoyeauP YosefN JordanMI RegierJ.Decision\u2010making with auto\u2010encoding variational Bayes. In:NIPS'20: Proceedings of the 34th International Conference on Neural Information Processing Systems.2020."},{"key":"e_1_2_9_20_1","doi-asserted-by":"crossref","unstructured":"AkcayS Atapour\u2010AbarghoueiA BreckonTP.GANomaly: semi\u2010supervised anomaly detection via adversarial training. In:Computer Vision \u2014ACCV 2018 Springer;2019:622\u2010637.","DOI":"10.1007\/978-3-030-20893-6_39"},{"key":"e_1_2_9_21_1","unstructured":"HoJ JainA AbbeelP.Denoising diffusion probabilistic models. In:NIPS '20: Proceedings of the 34th International Conference on Neural Information Processing Systems.Curran Associates Inc.;2020."},{"key":"e_1_2_9_22_1","unstructured":"DhariwalP NicholA.Diffusion models beat GANs on image synthesis. In:NIPS'21: Proceedings of the 35th International Conference on Neural Information Processing Systems.Curran Associates Inc.;2024."},{"key":"e_1_2_9_23_1","doi-asserted-by":"crossref","unstructured":"WyattJ LeachA SchmonSM WillcocksCG.AnoDDPM: anomaly detection with denoising diffusion probabilistic models using simplex noise. In:2022 IEEE\/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW).IEEE;2022:649\u2010655","DOI":"10.1109\/CVPRW56347.2022.00080"},{"key":"e_1_2_9_24_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.media.2023.102963"},{"key":"e_1_2_9_25_1","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-18576-2_4"},{"key":"e_1_2_9_26_1","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-16452-1_4"},{"key":"e_1_2_9_27_1","unstructured":"IqbalH KhalidU ChenC HuaJ.Unsupervised anomaly detection in medical images using masked diffusion model. In:Machine Learning in Medical Imaging: 14th International Workshop MLMI 2023 Held in Conjunction with MIC642 CAI 2023 Vancouver BC Canada October 8 2023 Proceedings Part I.Springer;2023."},{"key":"e_1_2_9_28_1","unstructured":"BerceaCI NeumayrM RueckertD SchnabelJA.Mask stitch and re\u2010sample: enhancing robustness and generalizability in anomaly detection through automatic diffusion models. inICML 3rd Workshop on Interpretable Machine Learning in Healthcare (IMLH) 2023."},{"key":"e_1_2_9_29_1","unstructured":"BehrendtF BhattacharyaD KrugerJ OpferR SchlaeferA.Patched diffusion models for unsupervised anomaly detection in brain MRI. In:Brain MRI in International Conference on Medical Imaging with Deep Learning.PMLR;2023."},{"key":"e_1_2_9_30_1","doi-asserted-by":"publisher","DOI":"10.3390\/data5010014"},{"key":"e_1_2_9_31_1","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-45673-2_15"},{"key":"e_1_2_9_32_1","doi-asserted-by":"publisher","DOI":"10.3389\/fnimg.2023.1157565"},{"key":"e_1_2_9_33_1","doi-asserted-by":"crossref","unstructured":"PinayaWHL GrahamMS GrayR et\u00a0al.Fast Unsupervised brain anomaly detection and segmentation with diffusion models. In:Medical Image Computing and Computer Assisted Intervention \u2014MICCAI 2022.Springer;2022:705\u2010714.","DOI":"10.1007\/978-3-031-16452-1_67"},{"key":"e_1_2_9_34_1","doi-asserted-by":"publisher","DOI":"10.1088\/0031-9155\/45\/2\/314"},{"key":"e_1_2_9_35_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.clineuro.2022.107258"},{"key":"e_1_2_9_36_1","unstructured":"ChungH KimJ McCannMT KlaskyML YeJC.Diffusion posterior sampling for general noisy inverse problems. inThe Eleventh International Conference on Learning Representations 2023."},{"key":"e_1_2_9_37_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.neuroimage.2008.10.040"},{"key":"e_1_2_9_38_1","unstructured":"KascenasA PugeaultN O'NeilAQ.Denoising autoencoders for unsupervised anomaly detection in brain MRI. In:Proceedings of The 5th International Conference on Medical Imaging with Deep Learning. Volume172PLMR;2022:653\u2010664."},{"key":"e_1_2_9_39_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.media.2019.01.010"},{"key":"e_1_2_9_40_1","first-page":"289","volume-title":"Medical Image Computing and Computer Assisted Intervention \u2014MICCAI 2019","author":"Zimmerer D","year":"2019"},{"key":"e_1_2_9_41_1","doi-asserted-by":"publisher","DOI":"10.1038\/s41592-020-01008-z"}],"container-title":["Medical Physics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/aapm.onlinelibrary.wiley.com\/doi\/pdf\/10.1002\/mp.17811","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,11,12]],"date-time":"2025-11-12T08:03:58Z","timestamp":1762934638000},"score":1,"resource":{"primary":{"URL":"https:\/\/aapm.onlinelibrary.wiley.com\/doi\/10.1002\/mp.17811"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,4,9]]},"references-count":40,"journal-issue":{"issue":"6","published-print":{"date-parts":[[2025,6]]}},"alternative-id":["10.1002\/mp.17811"],"URL":"https:\/\/doi.org\/10.1002\/mp.17811","archive":["Portico"],"relation":{},"ISSN":["0094-2405","2473-4209"],"issn-type":[{"value":"0094-2405","type":"print"},{"value":"2473-4209","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,4,9]]},"assertion":[{"value":"2024-11-04","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2025-03-10","order":2,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2025-04-09","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}