{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,15]],"date-time":"2026-10-15T21:37:47Z","timestamp":1792100267105,"version":"4.4.0"},"reference-count":44,"publisher":"Wiley","issue":"7","license":[{"start":{"date-parts":[[2025,5,30]],"date-time":"2025-05-30T00:00:00Z","timestamp":1748563200000},"content-version":"am","delay-in-days":365,"URL":"http:\/\/onlinelibrary.wiley.com\/termsAndConditions#am"},{"start":{"date-parts":[[2024,5,30]],"date-time":"2024-05-30T00:00:00Z","timestamp":1717027200000},"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":["#2138259"],"award-info":[{"award-number":["#2138259"]}],"id":[{"id":"10.13039\/100000001","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100000001","name":"National Science Foundation","doi-asserted-by":"publisher","award":["#2138286"],"award-info":[{"award-number":["#2138286"]}],"id":[{"id":"10.13039\/100000001","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100000001","name":"National Science Foundation","doi-asserted-by":"publisher","award":["#2138307"],"award-info":[{"award-number":["#2138307"]}],"id":[{"id":"10.13039\/100000001","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100000001","name":"National Science Foundation","doi-asserted-by":"publisher","award":["#2137603"],"award-info":[{"award-number":["#2137603"]}],"id":[{"id":"10.13039\/100000001","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100000001","name":"National Science Foundation","doi-asserted-by":"publisher","award":["#2138296"],"award-info":[{"award-number":["#2138296"]}],"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":[[2024,7]]},"abstract":"<jats:title>Abstract<\/jats:title><jats:sec><jats:title>Background<\/jats:title><jats:p>3D neural network dose predictions are useful for automating brachytherapy (BT) treatment planning for cervical cancer. Cervical BT can be delivered with numerous applicators, which necessitates developing models that generalize to multiple applicator types. The variability and scarcity of data for any given applicator type poses challenges for deep learning.<\/jats:p><\/jats:sec><jats:sec><jats:title>Purpose<\/jats:title><jats:p>The goal of this work was to compare three methods of neural network training\u2014a single model trained on all applicator data, fine\u2010tuning the combined model to each applicator, and individual (IDV) applicator models\u2014to determine the optimal method for dose prediction.<\/jats:p><\/jats:sec><jats:sec><jats:title>Methods<\/jats:title><jats:p>Models were produced for four applicator types\u2014tandem\u2010and\u2010ovoid (T&amp;O), T&amp;O with 1\u20137 needles (T&amp;ON), tandem\u2010and\u2010ring (T&amp;R) and T&amp;R with 1\u20134 needles (T&amp;RN). First, the combined model was trained on 859 treatment plans from 266 cervical cancer patients treated from 2010 onwards. The train\/validation\/test split was 70%\/16%\/14%, with approximately 49%\/10%\/19%\/22% T&amp;O\/T&amp;ON\/T&amp;R\/T&amp;RN in each dataset. Inputs included four channels for anatomical masks (high\u2010risk clinical target volume [HRCTV], bladder, rectum, and sigmoid), a mask indicating dwell position locations, and applicator channels for each applicator component. Applicator channels were created by mapping the 3D dose for a single dwell position to each dwell position and summing over each applicator component with uniform dwell time weighting. A 3D Cascade U\u2010Net, which consists of two U\u2010Nets in sequence, and mean squared error loss function were used. The combined model was then fine\u2010tuned to produce four applicator\u2010specific models by freezing the first U\u2010Net and encoding layers of the second and resuming training on applicator\u2010specific data. Finally, four IDV models were trained using only data from each applicator type. Performance of these three model types was compared using the following metrics for the test set: mean error (ME, representing model bias) and mean absolute error (MAE) over all dose voxels and ME of clinical metrics (HRCTV D90% and <jats:italic>D<\/jats:italic><jats:sub>2cc<\/jats:sub> of bladder, rectum, and sigmoid), averaged over all patients. A positive ME indicates the clinical dose was higher than predicted. 3D global gamma analysis with the prescription dose as reference value was performed. Dice similarity coefficients (DSC) were computed for each isodose volume.<\/jats:p><\/jats:sec><jats:sec><jats:title>Results<\/jats:title><jats:p>Fine\u2010tuned and combined models showed better performance than IDV applicator training. Fine\u2010tuning resulted in modest improvements in about half the metrics, compared to the combined model, while the remainder were mostly unchanged. Fine\u2010tuned MAE\u00a0=\u00a03.98%\/2.69%\/5.36%\/3.80% for T&amp;O\/T&amp;R\/T&amp;ON\/T&amp;RN, and ME over all voxels\u00a0=\u00a0\u20130.08%\/\u20130.89%\/\u20130.59%\/1.42%. ME <jats:italic>D<\/jats:italic><jats:sub>2cc<\/jats:sub> were bladder\u00a0=\u00a0\u20130.77%\/1.00%\/\u20130.66%\/\u20131.53%, rectum\u00a0=\u00a01.11%\/\u20130.22%\/\u20130.29%\/\u20133.37%, sigmoid\u00a0=\u00a0\u20130.47%\/\u20130.06%\/\u20132.37%\/\u20131.40%, and ME D90\u00a0=\u00a02.6%\/\u20134.4%\/4.8%\/0.0%. Gamma pass rates (3%\/3\u00a0mm) were 86%\/91%\/83%\/89%. Mean DSCs were 0.92%\/0.92%\/0.88%\/0.91% for isodoses \u2264\u00a0150% of prescription.<\/jats:p><\/jats:sec><jats:sec><jats:title>Conclusions<\/jats:title><jats:p>3D BT dose was accurately predicted for all applicator types, as indicated by the low MAE and MEs, high gamma scores and high DSCs. Training on all treatment data overcomes challenges with data scarcity in each applicator type, resulting in superior performance than can be achieved by training on IDV applicators alone. This could presumably be explained by the fact that the larger, more diverse dataset allows the neural network to learn underlying trends and characteristics in dose that are common to all treatment applicators. Accurate, applicator\u2010specific dose predictions could enable automated, knowledge\u2010based planning for any cervical brachytherapy treatment.<\/jats:p><\/jats:sec>","DOI":"10.1002\/mp.17230","type":"journal-article","created":{"date-parts":[[2024,5,30]],"date-time":"2024-05-30T13:34:00Z","timestamp":1717076040000},"page":"4591-4606","update-policy":"https:\/\/doi.org\/10.1002\/crossmark_policy","source":"Crossref","is-referenced-by-count":10,"title":["Neural network dose prediction for cervical brachytherapy: Overcoming data scarcity for applicator\u2010specific models"],"prefix":"10.1002","volume":"51","author":[{"given":"Lance C.","family":"Moore","sequence":"first","affiliation":[{"name":"Radiation Medicine and Applied Sciences University of California San Diego  La Jolla California USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Fritz","family":"Ahern","sequence":"additional","affiliation":[{"name":"Radiation Medicine and Applied Sciences University of California San Diego  La Jolla California USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Lingyi","family":"Li","sequence":"additional","affiliation":[{"name":"Radiation Medicine and Applied Sciences University of California San Diego  La Jolla California USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Karoline","family":"Kallis","sequence":"additional","affiliation":[{"name":"Radiation Medicine and Applied Sciences University of California San Diego  La Jolla California USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Kelly","family":"Kisling","sequence":"additional","affiliation":[{"name":"Radiation Medicine and Applied Sciences University of California San Diego  La Jolla California USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Katherina G.","family":"Cortes","sequence":"additional","affiliation":[{"name":"Radiation Medicine and Applied Sciences University of California San Diego  La Jolla California USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Chika","family":"Nwachukwu","sequence":"additional","affiliation":[{"name":"Radiation Medicine and Applied Sciences University of California San Diego  La Jolla California USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Dominique","family":"Rash","sequence":"additional","affiliation":[{"name":"Radiation Medicine and Applied Sciences University of California San Diego  La Jolla California USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Catheryn M.","family":"Yashar","sequence":"additional","affiliation":[{"name":"Radiation Medicine and Applied Sciences University of California San Diego  La Jolla California USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jyoti","family":"Mayadev","sequence":"additional","affiliation":[{"name":"Radiation Medicine and Applied Sciences University of California San Diego  La Jolla California USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jingjing","family":"Zou","sequence":"additional","affiliation":[{"name":"Herbert Wertheim School of Public Health and Human Longevity Science University of California San Diego and Moores Cancer Center  La Jolla California USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Nuno","family":"Vasconcelos","sequence":"additional","affiliation":[{"name":"Electrical and Computer Engineering University of California San Diego  La Jolla California USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sandra M.","family":"Meyers","sequence":"additional","affiliation":[{"name":"Radiation Medicine and Applied Sciences University of California San Diego  La Jolla California USA"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"311","published-online":{"date-parts":[[2024,5,30]]},"reference":[{"key":"e_1_2_8_2_1","doi-asserted-by":"publisher","DOI":"10.1118\/1.3253464"},{"key":"e_1_2_8_3_1","doi-asserted-by":"publisher","DOI":"10.1118\/1.3539749"},{"key":"e_1_2_8_4_1","doi-asserted-by":"publisher","DOI":"10.1118\/1.4761864"},{"key":"e_1_2_8_5_1","doi-asserted-by":"publisher","DOI":"10.1002\/mp.13526"},{"key":"e_1_2_8_6_1","doi-asserted-by":"publisher","DOI":"10.1002\/mp.13896"},{"key":"e_1_2_8_7_1","doi-asserted-by":"publisher","DOI":"10.1088\/1361\u20106560\/acc37c"},{"key":"e_1_2_8_8_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.ijrobp.2021.06.028"},{"key":"e_1_2_8_9_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.ijrobp.2019.10.036"},{"key":"e_1_2_8_10_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.brachy.2016.06.001"},{"key":"e_1_2_8_11_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.brachy.2014.01.004"},{"key":"e_1_2_8_12_1","doi-asserted-by":"publisher","DOI":"10.1120\/jacmp.v16i2.5204"},{"key":"e_1_2_8_13_1","doi-asserted-by":"publisher","DOI":"10.1186\/s13014\u2010015\u20100542\u20101"},{"key":"e_1_2_8_14_1","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0145137"},{"key":"e_1_2_8_15_1","doi-asserted-by":"publisher","DOI":"10.1118\/1.4938583"},{"key":"e_1_2_8_16_1","doi-asserted-by":"publisher","DOI":"10.1088\/1361\u20106560\/ab039b"},{"key":"e_1_2_8_17_1","doi-asserted-by":"publisher","DOI":"10.1002\/mp.13955"},{"key":"e_1_2_8_18_1","doi-asserted-by":"publisher","DOI":"10.1002\/mp.14845"},{"key":"e_1_2_8_19_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.brachy.2022.03.002"},{"key":"e_1_2_8_20_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.ijrobp.2020.04.045"},{"key":"e_1_2_8_21_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.compbiomed.2021.104755"},{"key":"e_1_2_8_22_1","doi-asserted-by":"publisher","DOI":"10.1109\/TRPMS.2021.3098507"},{"key":"e_1_2_8_23_1","doi-asserted-by":"publisher","DOI":"10.1088\/1361\u20106560\/acecd2"},{"key":"e_1_2_8_24_1","doi-asserted-by":"publisher","DOI":"10.1002\/acm2.13942"},{"key":"e_1_2_8_25_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.brachy.2011.07.003"},{"key":"e_1_2_8_26_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.ctro.2018.01.001"},{"key":"e_1_2_8_27_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.prro.2021.02.003"},{"key":"e_1_2_8_28_1","doi-asserted-by":"publisher","DOI":"10.1002\/mp.15034"},{"key":"e_1_2_8_29_1","doi-asserted-by":"crossref","unstructured":"CholletF.Xception: deep learning with depthwise separable convolutions. In:2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).IEEE;2017:1800\u20101807. doi:10.48550\/arXiv.1610.02357","DOI":"10.1109\/CVPR.2017.195"},{"key":"e_1_2_8_30_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.cmpb.2021.106236"},{"key":"e_1_2_8_31_1","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1412.6980"},{"key":"e_1_2_8_32_1","doi-asserted-by":"crossref","unstructured":"HeK ZhangX RenS SunJ.Delving deep into rectifiers: surpassing human\u2010level performance on ImageNet classification. In:2015 IEEE International Conference on Computer Vision (ICCV).IEEE;2015:1026\u20101034. doi:10.48550\/arXiv.1502.01852","DOI":"10.1109\/ICCV.2015.123"},{"key":"e_1_2_8_33_1","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-92185-9_46"},{"key":"e_1_2_8_34_1","unstructured":"LeeCY XieS GallagherP ZhangZ TuZ.Deeply\u2010supervised nets. In:Proceedings of the Eighteenth International Conference on Artificial Intelligence and Statistics.PMLR;2015:562\u2010570.https:\/\/proceedings.mlr.press\/v38\/lee15a.html"},{"key":"e_1_2_8_35_1","doi-asserted-by":"crossref","unstructured":"SmithLN.Cyclical learning rates for training neural networks. In:2017 IEEE Winter Conference on Applications of Computer Vision (WACV).IEEE;2017:464\u2010472. doi:10.48550\/arXiv.1506.01186","DOI":"10.1109\/WACV.2017.58"},{"key":"e_1_2_8_36_1","doi-asserted-by":"publisher","DOI":"10.1177\/00368504211038162"},{"key":"e_1_2_8_37_1","doi-asserted-by":"publisher","DOI":"10.1186\/s13014\u2010021\u201001864\u20109"},{"key":"e_1_2_8_38_1","doi-asserted-by":"publisher","DOI":"10.1002\/mp.15461"},{"key":"e_1_2_8_39_1","doi-asserted-by":"publisher","DOI":"10.1002\/mp.13490"},{"key":"e_1_2_8_40_1","doi-asserted-by":"publisher","DOI":"10.1002\/mp.13262"},{"key":"e_1_2_8_41_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.prro.2011.11.012"},{"key":"e_1_2_8_42_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.ijrobp.2015.01.046"},{"key":"e_1_2_8_43_1","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0232697"},{"key":"e_1_2_8_44_1","volume-title":"Advances in Neural Information Processing Systems","author":"Rebuffi SA","year":"2023"},{"key":"e_1_2_8_45_1","unstructured":"IqbalH.PlotNeuralNet. Published online September 3 2023. Accessed September 3 2023.https:\/\/github.com\/HarisIqbal88\/PlotNeuralNet"}],"container-title":["Medical Physics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/aapm.onlinelibrary.wiley.com\/doi\/am-pdf\/10.1002\/mp.17230","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"https:\/\/aapm.onlinelibrary.wiley.com\/doi\/pdf\/10.1002\/mp.17230","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,7,9]],"date-time":"2024-07-09T15:52:54Z","timestamp":1720540374000},"score":1,"resource":{"primary":{"URL":"https:\/\/aapm.onlinelibrary.wiley.com\/doi\/10.1002\/mp.17230"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,5,30]]},"references-count":44,"journal-issue":{"issue":"7","published-print":{"date-parts":[[2024,7]]}},"alternative-id":["10.1002\/mp.17230"],"URL":"https:\/\/doi.org\/10.1002\/mp.17230","archive":["Portico"],"relation":{},"ISSN":["0094-2405","2473-4209"],"issn-type":[{"value":"0094-2405","type":"print"},{"value":"2473-4209","type":"electronic"}],"subject":[],"published":{"date-parts":[[2024,5,30]]},"assertion":[{"value":"2023-11-16","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2024-05-11","order":1,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2024-05-30","order":2,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}