{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,11,5]],"date-time":"2026-11-05T04:50:59Z","timestamp":1793854259571,"version":"4.6.1"},"reference-count":41,"publisher":"SPIE-Intl Soc Optical Eng","issue":"01","license":[{"start":{"date-parts":[[2018,10,1]],"date-time":"2018-10-01T00:00:00Z","timestamp":1538352000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100003185","name":"Fraunhofer-Gesellschaft","doi-asserted-by":"publisher","id":[{"id":"10.13039\/501100003185","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100007210","name":"Varian Medical Systems, Inc.","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100007210","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["J. Med. Imag."],"published-print":{"date-parts":[[2018,10,1]]},"abstract":"<jats:p>The segmentation of organs at risk is a crucial and time-consuming step in radiotherapy planning. Good automatic methods can significantly reduce the time clinicians have to spend on this task. Due to its variability in shape and low contrast to surrounding structures, segmenting the parotid gland is challenging. Motivated by the recent success of deep learning, we study the use of two-dimensional (2-D), 2-D ensemble, and three-dimensional (3-D) U-Nets for segmentation. The mean Dice similarity to ground truth is \u223c0.83 for all three models. A patch-based approach for class balancing seems promising for false-positive reduction. The 2-D ensemble and 3-D U-Net are applied to the test data of the 2015 MICCAI challenge on head and neck autosegmentation. Both deep learning methods generalize well onto independent data (Dice 0.865 and 0.88) and are superior to a selection of model- and atlas-based methods with respect to the Dice coefficient. Since appropriate reference annotations are essential for training but often difficult and expensive to obtain, it is important to know how many samples are needed for training. We evaluate the performance after training with different-sized training sets and observe no significant increase in the Dice coefficient for more than 250 training cases.<\/jats:p>","DOI":"10.1117\/1.jmi.6.1.011005","type":"journal-article","created":{"date-parts":[[2018,10,1]],"date-time":"2018-10-01T00:11:45Z","timestamp":1538352705000},"page":"1-9","source":"Crossref","is-referenced-by-count":24,"title":["Evaluation of deep learning methods for parotid gland segmentation from CT images"],"prefix":"10.1117","volume":"6","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-4327-6352","authenticated-orcid":false,"given":"Annika","family":"H\u00e4nsch","sequence":"first","affiliation":[{"name":"Fraunhofer MEVIS"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Michael","family":"Schwier","sequence":"first","affiliation":[{"name":"Fraunhofer MEVIS"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Tobias","family":"Gass","sequence":"first","affiliation":[{"name":"Varian Medical Systems, Inc."}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Tomasz","family":"Morgas","sequence":"first","affiliation":[{"name":"Varian Medical Systems, Inc."}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Benjamin","family":"Haas","sequence":"first","affiliation":[{"name":"Varian Medical Systems, Inc."}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Volker","family":"Dicken","sequence":"additional","affiliation":[{"name":"Fraunhofer MEVIS"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-7557-5007","authenticated-orcid":false,"given":"Hans","family":"Meine","sequence":"additional","affiliation":[{"name":"Fraunhofer MEVIS"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-9421-3549","authenticated-orcid":false,"given":"Jan","family":"Klein","sequence":"additional","affiliation":[{"name":"Fraunhofer MEVIS"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Horst K.","family":"Hahn","sequence":"additional","affiliation":[{"name":"Fraunhofer MEVIS"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"189","reference":[{"key":"r1","doi-asserted-by":"publisher","DOI":"10.1177\/0022034509343143"},{"key":"r2","doi-asserted-by":"publisher","DOI":"10.2147\/CMAR.S47220"},{"key":"r3","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-540-85990-1"},{"key":"r4","doi-asserted-by":"publisher","DOI":"10.1118\/1.4810971"},{"key":"r5","doi-asserted-by":"publisher","DOI":"10.1109\/TMI.2015.2442753"},{"key":"r6","article-title":"A multi-atlas approach for the automatic segmentation of multiple structures in head and neck CT images","author":"Chen","year":"2015"},{"key":"r7","doi-asserted-by":"publisher","DOI":"10.1118\/1.3654160"},{"key":"r8","article-title":"Automatic segmentation of structures in CT head and neck images using a coupled shape model","author":"Jung","year":"2015"},{"key":"r9","article-title":"Fully automatic segmentation of head and neck organs using active appearance models","author":"Mannion-Haworth","year":"2015"},{"key":"r10","article-title":"Head and neck auto segmentation challenge based on non-local generative models","author":"Orbes-Arteaga","year":"2015"},{"key":"r11","doi-asserted-by":"crossref","DOI":"10.54294\/kmcunc","article-title":"Multi atlas segmentation with active shape model refinement for multi-organ segmentation in head and neck cancer radiotherapy planning","author":"Albrecht","year":"2015"},{"key":"r12","doi-asserted-by":"publisher","DOI":"10.1118\/1.4871623"},{"key":"r13","doi-asserted-by":"publisher","DOI":"10.1002\/mp.12197"},{"key":"r14","doi-asserted-by":"publisher","DOI":"10.1016\/j.media.2017.07.005"},{"key":"r15","doi-asserted-by":"publisher","DOI":"10.1146\/annurev-bioeng-071516-044442"},{"key":"r16","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-24574-4"},{"key":"r17","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-46723-8"},{"key":"r18","doi-asserted-by":"publisher","DOI":"10.1109\/3DV.2016.79"},{"key":"r19","doi-asserted-by":"publisher","DOI":"10.1016\/j.media.2016.10.004"},{"key":"r20","doi-asserted-by":"publisher","DOI":"10.1117\/12.2293528"},{"key":"r21","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-46976-8"},{"key":"r22","doi-asserted-by":"publisher","DOI":"10.1007\/s12021-018-9377-x"},{"key":"r23","article-title":"Multimodal brain tumor segmentation (BraTS), Challenge 2017","year":"2018"},{"key":"r24","article-title":"Liver tumor segmentation (LiTS), Challenge 2017","year":"2018"},{"key":"r25","article-title":"Ischemic stroke lesion segmentation (ISLES), Challenge 2017","year":"2018"},{"key":"r26","doi-asserted-by":"publisher","DOI":"10.1016\/j.compbiomed.2018.05.018"},{"key":"r27","doi-asserted-by":"publisher","DOI":"10.1016\/j.radonc.2018.05.030"},{"key":"r28","doi-asserted-by":"publisher","DOI":"10.1016\/j.radonc.2017.11.012"},{"key":"r29","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-46723-8"},{"key":"r30","doi-asserted-by":"publisher","DOI":"10.1002\/mp.12045"},{"key":"r31","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2009.5206848"},{"key":"r32","article-title":"PDDCA: a public domain database for computational anatomy","author":"Sharp","year":"2016"},{"key":"r33","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-642-40763-5"},{"key":"r34","article-title":"Lasagne: first release","author":"Dieleman","year":"2015"},{"key":"r35","first-page":"448","article-title":"Batch normalization: accelerating deep network training by reducing internal covariate shift","author":"Ioffe","year":"2015"},{"key":"r36","first-page":"4898","article-title":"Understanding the effective receptive field in deep convolutional neural networks","author":"Luo","year":"2016"},{"key":"r37","doi-asserted-by":"publisher","DOI":"10.2307\/1932409"},{"key":"r38","doi-asserted-by":"publisher","DOI":"10.1109\/MPUL.2011.942929"},{"key":"r39","doi-asserted-by":"publisher","DOI":"10.1109\/TMI.2009.2013851"},{"key":"r40","doi-asserted-by":"publisher","DOI":"10.1016\/S0167-8140(18)30840-5"},{"key":"r41","doi-asserted-by":"publisher","DOI":"10.1117\/12.2292962"}],"container-title":["Journal of Medical Imaging"],"original-title":[],"link":[{"URL":"https:\/\/www.spiedigitallibrary.org\/journals\/journal-of-medical-imaging\/volume-6\/issue-01\/011005\/Evaluation-of-deep-learning-methods-for-parotid-gland-segmentation-from\/10.1117\/1.JMI.6.1.011005.full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,25]],"date-time":"2026-09-25T00:31:11Z","timestamp":1790296271000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.spiedigitallibrary.org\/journals\/journal-of-medical-imaging\/volume-6\/issue-01\/011005\/Evaluation-of-deep-learning-methods-for-parotid-gland-segmentation-from\/10.1117\/1.JMI.6.1.011005.full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2018,10,1]]},"references-count":41,"journal-issue":{"issue":"01"},"URL":"https:\/\/doi.org\/10.1117\/1.jmi.6.1.011005","relation":{},"ISSN":["2329-4302","2329-4310"],"issn-type":[{"value":"2329-4302","type":"print"},{"value":"2329-4310","type":"electronic"}],"subject":[],"published":{"date-parts":[[2018,10,1]]}}}