{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,1]],"date-time":"2026-08-01T15:28:13Z","timestamp":1785598093097,"version":"3.56.0"},"reference-count":43,"publisher":"Wiley","issue":"3","license":[{"start":{"date-parts":[[2023,9,11]],"date-time":"2023-09-11T00:00:00Z","timestamp":1694390400000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/onlinelibrary.wiley.com\/termsAndConditions#vor"}],"funder":[{"DOI":"10.13039\/100000001","name":"National Science Foundation","doi-asserted-by":"publisher","award":["CA014236"],"award-info":[{"award-number":["CA014236"]}],"id":[{"id":"10.13039\/100000001","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":[[2024,3]]},"abstract":"<jats:title>Abstract<\/jats:title><jats:sec><jats:title>Background<\/jats:title><jats:p>Uncertainty quantification in deep learning is an important research topic. For medical image segmentation, the uncertainty measurements are usually reported as the likelihood that each pixel belongs to the predicted segmentation region. In potential clinical applications, the uncertainty result reflects the algorithm's robustness and supports the confidence and trust of the segmentation result when the ground\u2010truth result is absent. For commonly studied deep learning models, novel methods for quantifying segmentation uncertainty are in demand.<\/jats:p><\/jats:sec><jats:sec><jats:title>Purpose<\/jats:title><jats:p>To develop a U\u2010Net segmentation uncertainty quantification method based on spherical image projection of multi\u2010parametric MRI (MP\u2010MRI) in glioma segmentation.<\/jats:p><\/jats:sec><jats:sec><jats:title>Methods<\/jats:title><jats:p>The projection of planar MRI data onto a spherical surface is equivalent to a nonlinear image transformation that retains global anatomical information. By incorporating this image transformation process in our proposed spherical projection\u2010based U\u2010Net (SPU\u2010Net) segmentation model design, multiple independent segmentation predictions can be obtained from a single MRI. The final segmentation is the average of all available results, and the variation can be visualized as a pixel\u2010wise uncertainty map. An uncertainty score was introduced to evaluate and compare the performance of uncertainty measurements.<\/jats:p><jats:p>The proposed SPU\u2010Net model was implemented on the basis of 369 glioma patients with MP\u2010MRI scans (T1, T1\u2010Ce, T2, and FLAIR). Three SPU\u2010Net models were trained to segment enhancing tumor (ET), tumor core (TC), and whole tumor (WT), respectively. The SPU\u2010Net model was compared with (1) the classic U\u2010Net model with test\u2010time augmentation (TTA) and (2) linear scaling\u2010based U\u2010Net (LSU\u2010Net) segmentation models in terms of both segmentation accuracy (Dice coefficient, sensitivity, specificity, and accuracy) and segmentation uncertainty (uncertainty map and uncertainty score).<\/jats:p><\/jats:sec><jats:sec><jats:title>Results<\/jats:title><jats:p>The developed SPU\u2010Net model successfully achieved low uncertainty for correct segmentation predictions (e.g., tumor interior or healthy tissue interior) and high uncertainty for incorrect results (e.g., tumor boundaries). This model could allow the identification of missed tumor targets or segmentation errors in U\u2010Net. Quantitatively, the SPU\u2010Net model achieved the highest uncertainty scores for three segmentation targets (ET\/TC\/WT): 0.826\/0.848\/0.936, compared to 0.784\/0.643\/0.872 using the U\u2010Net with TTA and 0.743\/0.702\/0.876 with the LSU\u2010Net (scaling factor\u00a0=\u00a02). The SPU\u2010Net also achieved statistically significantly higher Dice coefficients, underscoring the improved segmentation accuracy.<\/jats:p><\/jats:sec><jats:sec><jats:title>Conclusion<\/jats:title><jats:p>The SPU\u2010Net model offers a powerful tool to quantify glioma segmentation uncertainty while improving segmentation accuracy. The proposed method can be generalized to other medical image\u2010related deep\u2010learning applications for uncertainty evaluation.<\/jats:p><\/jats:sec>","DOI":"10.1002\/mp.16695","type":"journal-article","created":{"date-parts":[[2023,9,11]],"date-time":"2023-09-11T20:59:16Z","timestamp":1694465956000},"page":"1931-1943","update-policy":"https:\/\/doi.org\/10.1002\/crossmark_policy","source":"Crossref","is-referenced-by-count":12,"title":["Quantifying U\u2010Net uncertainty in multi\u2010parametric MRI\u2010based glioma segmentation by spherical image projection"],"prefix":"10.1002","volume":"51","author":[{"given":"Zhenyu","family":"Yang","sequence":"first","affiliation":[{"name":"Department of Radiation Oncology Duke University  Durham North Carolina USA"},{"name":"Medical Physics Graduate Program Duke Kunshan University  Kunshan Jiangsu China"},{"name":"Medical Physics Graduate Program Duke University  Durham North Carolina USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Kyle","family":"Lafata","sequence":"additional","affiliation":[{"name":"Department of Radiation Oncology Duke University  Durham North Carolina USA"},{"name":"Department of Radiology Duke University  Durham North Carolina USA"},{"name":"Department of Electrical and Computer Engineering Duke University  Durham North Carolina USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Eugene","family":"Vaios","sequence":"additional","affiliation":[{"name":"Department of Radiation Oncology Duke University  Durham North Carolina USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Zongsheng","family":"Hu","sequence":"additional","affiliation":[{"name":"Department of Radiation Physics The University of Texas MD Anderson Cancer Center  Houston Texas USA"},{"name":"The University of Texas MD Anderson Graduate School of Biomedical Science  Houston Texas USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Trey","family":"Mullikin","sequence":"additional","affiliation":[{"name":"Department of Radiation Oncology Duke University  Durham North Carolina USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Fang\u2010Fang","family":"Yin","sequence":"additional","affiliation":[{"name":"Department of Radiation Oncology Duke University  Durham North Carolina USA"},{"name":"Medical Physics Graduate Program Duke Kunshan University  Kunshan Jiangsu China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Chunhao","family":"Wang","sequence":"additional","affiliation":[{"name":"Department of Radiation Oncology Duke University  Durham North Carolina USA"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"311","published-online":{"date-parts":[[2023,9,11]]},"reference":[{"key":"e_1_2_8_2_1","doi-asserted-by":"publisher","DOI":"10.1146\/annurev.bioeng.2.1.315"},{"key":"e_1_2_8_3_1","article-title":"Brain tumor segmentation with self\u2010ensembled, deeply\u2010supervised 3D U\u2010net neural networks: a BraTS 2020 challenge solution","author":"Henry T","year":"2020","journal-title":"ArXiv201101045 Cs Eess"},{"key":"e_1_2_8_4_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.procs.2016.09.407"},{"key":"e_1_2_8_5_1","doi-asserted-by":"publisher","DOI":"10.1007\/s10278-019-00227-x"},{"key":"e_1_2_8_6_1","doi-asserted-by":"publisher","DOI":"10.1007\/s40747-022-00815-5"},{"key":"e_1_2_8_7_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.inffus.2021.05.008"},{"key":"e_1_2_8_8_1","unstructured":"GawlikowskiJ TassiCRN AliM et\u00a0al.A survey of uncertainty in deep neural networks. Published online January 18 2022. Accessed July 25 2022.http:\/\/arxiv.org\/abs\/2107.03342"},{"key":"e_1_2_8_9_1","article-title":"A neural ordinary differential equation model for visualizing deep neural network behaviors in multi\u2010parametric MRI based glioma segmentation","author":"Yang Z","year":"2022","journal-title":"ArXiv Prepr ArXiv220300628"},{"key":"e_1_2_8_10_1","doi-asserted-by":"crossref","unstructured":"SanchezT CaramiauxB ThielP MackayWE.Deep learning uncertainty in machine teaching. In:27th International Conference on Intelligent User Interfaces.ACM;2022:173\u2010190. doi:10.1145\/3490099.3511117","DOI":"10.1145\/3490099.3511117"},{"key":"e_1_2_8_11_1","doi-asserted-by":"publisher","DOI":"10.1111\/coin.12411"},{"key":"e_1_2_8_12_1","first-page":"682","volume-title":"On the effect of inter\u2010observer variability for a reliable estimation of uncertainty of medical image segmentation","author":"Jungo A","year":"2018"},{"key":"e_1_2_8_13_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.csda.2019.106816"},{"key":"e_1_2_8_14_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.strusafe.2008.06.020"},{"key":"e_1_2_8_15_1","doi-asserted-by":"publisher","DOI":"10.3390\/rs13193836"},{"key":"e_1_2_8_16_1","first-page":"1050","volume-title":"Dropout as a bayesian approximation: Representing model uncertainty in deep learning","author":"Gal Y","year":"2016"},{"key":"e_1_2_8_17_1","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-60365-6_4"},{"key":"e_1_2_8_18_1","doi-asserted-by":"publisher","DOI":"10.2174\/157340507782446241"},{"key":"e_1_2_8_19_1","doi-asserted-by":"crossref","unstructured":"MundtM PliushchI MajumderS RameshV.Open set recognition through deep neural network uncertainty: does out\u2010of\u2010distribution detection require generative classifiers?In:2019 IEEE\/CVF International Conference on Computer Vision Workshop (ICCVW).IEEE;2019:753\u2010757. doi:10.1109\/ICCVW.2019.00098","DOI":"10.1109\/ICCVW.2019.00098"},{"key":"e_1_2_8_20_1","unstructured":"LiangS LiY SrikantR Enhancing The Reliability of Out\u2010of\u2010distribution Image Detection in Neural Networks. Published online August 30 2020. Accessed July 25 2022.http:\/\/arxiv.org\/abs\/1706.02690"},{"key":"e_1_2_8_21_1","doi-asserted-by":"publisher","DOI":"10.1186\/s40537\u2010019\u20100197\u20100"},{"key":"e_1_2_8_22_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.neucom.2019.01.103"},{"key":"e_1_2_8_23_1","first-page":"61","volume-title":"Automatic brain tumor segmentation using convolutional neural networks with test\u2010time augmentation","author":"Wang G","year":"2018"},{"key":"e_1_2_8_24_1","first-page":"234","volume-title":"U\u2010net: Convolutional networks for biomedical image segmentation","author":"Ronneberger O","year":"2015"},{"key":"e_1_2_8_25_1","doi-asserted-by":"publisher","DOI":"10.1109\/TMI.2014.2377694"},{"key":"e_1_2_8_26_1","first-page":"386","article-title":"Full\u2010view spherical image camera","volume":"4","author":"Li S","year":"2006","journal-title":"IEEE"},{"key":"e_1_2_8_27_1","unstructured":"ZhangC HeS LiwickiS.A spherical approach to planar semantic segmentation.2020."},{"key":"e_1_2_8_28_1","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2021.3086020"},{"key":"e_1_2_8_29_1","doi-asserted-by":"publisher","DOI":"10.2352\/J.ImagingSci.Technol.2020.64.2.020508"},{"key":"e_1_2_8_30_1","doi-asserted-by":"crossref","unstructured":"HenryT CarreA.LerousseauM et al.Brain tumor segmentation with self\u2010ensembled deeply\u2010supervised 3D U\u2010net neural networks: a BraTS 2020 challenge solution[C]\/\/Brainlesion: Glioma Multiple Sclerosis Stroke and Traumatic Brain Injuries: 6th International Workshop BrainLes 2020 Held in Conjunction with MICCAI 2020 Lima Peru October 4 2020 Revised Selected Papers Part I 6. Springer International Publishing 2021 327\u2010339.","DOI":"10.1007\/978-3-030-72084-1_30"},{"key":"e_1_2_8_31_1","article-title":"QU\u2010BraTS: mICCAI BraTS 2020 challenge on quantifying uncertainty in brain tumor segmentation\u2014analysis of ranking metrics and benchmarking results","author":"Mehta R","year":"2021","journal-title":"ArXiv211210074 Cs Eess"},{"key":"e_1_2_8_32_1","unstructured":"VoT.Attention! Stay Focus!Published online April 16 2021. Accessed September 14 2022.http:\/\/arxiv.org\/abs\/2104.07925"},{"key":"e_1_2_8_33_1","unstructured":"OktayO SchlemperJ FolgocLL et\u00a0al.Attention U\u2010Net: Learning Where to Look for the Pancreas. arXiv preprint arXiv:1804.03999 2018."},{"key":"e_1_2_8_34_1","first-page":"715","volume-title":"Is segmentation uncertainty useful?","author":"Czolbe S","year":"2021"},{"key":"e_1_2_8_35_1","volume-title":"Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support (pp. 3\u201011)","author":"Zhou Z"},{"key":"e_1_2_8_36_1","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2021.3122543"},{"key":"e_1_2_8_37_1","first-page":"156","article-title":"Uncertainty\u2010aware training of neural networks for selective medical image segmentation","author":"Ding Y","year":"2020","journal-title":"PMLR"},{"key":"e_1_2_8_38_1","doi-asserted-by":"publisher","DOI":"10.1007\/s00330-018-5695-5"},{"key":"e_1_2_8_39_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.media.2019.101557"},{"key":"e_1_2_8_40_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.neunet.2019.08.025"},{"key":"e_1_2_8_41_1","article-title":"Understanding the effective receptive field in deep convolutional neural networks","volume":"29","author":"Luo W","year":"2016","journal-title":"Adv Neural Inf Process Syst"},{"key":"e_1_2_8_42_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.media.2017.10.002"},{"key":"e_1_2_8_43_1","doi-asserted-by":"crossref","unstructured":"PengC ZhangX YuG et al.Large kernel matters\u2013improve semantic segmentation by global convolutional network[C]\/\/Proceedings of the IEEE conference on computer vision and pattern recognition.2017;4353\u20104361.","DOI":"10.1109\/CVPR.2017.189"},{"key":"e_1_2_8_44_1","doi-asserted-by":"crossref","unstructured":"ShenX WangC LiX et\u00a0al.Rf\u2010net: an end\u2010to\u2010end image matching network based on receptive field. Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition.2019;8132\u20108140.","DOI":"10.1109\/CVPR.2019.00832"}],"container-title":["Medical Physics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/aapm.onlinelibrary.wiley.com\/doi\/pdf\/10.1002\/mp.16695","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,3,12]],"date-time":"2024-03-12T14:59:22Z","timestamp":1710255562000},"score":1,"resource":{"primary":{"URL":"https:\/\/aapm.onlinelibrary.wiley.com\/doi\/10.1002\/mp.16695"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023,9,11]]},"references-count":43,"journal-issue":{"issue":"3","published-print":{"date-parts":[[2024,3]]}},"alternative-id":["10.1002\/mp.16695"],"URL":"https:\/\/doi.org\/10.1002\/mp.16695","archive":["Portico"],"relation":{},"ISSN":["0094-2405","2473-4209"],"issn-type":[{"value":"0094-2405","type":"print"},{"value":"2473-4209","type":"electronic"}],"subject":[],"published":{"date-parts":[[2023,9,11]]},"assertion":[{"value":"2022-11-08","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2023-08-08","order":1,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2023-09-11","order":2,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}