{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,17]],"date-time":"2026-10-17T05:17:49Z","timestamp":1792214269022,"version":"4.4.0"},"reference-count":38,"publisher":"Wiley","issue":"6","license":[{"start":{"date-parts":[[2021,4,25]],"date-time":"2021-04-25T00:00:00Z","timestamp":1619308800000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/onlinelibrary.wiley.com\/termsAndConditions#vor"}],"funder":[{"DOI":"10.13039\/100000002","name":"National Institutes of Health","doi-asserted-by":"publisher","award":["R01CA201212"],"award-info":[{"award-number":["R01CA201212"]}],"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":[[2021,6]]},"abstract":"<jats:sec>\n                    <jats:title>Purpose<\/jats:title>\n                    <jats:p>To develop an artificial intelligence (AI) agent for fully automated rapid head\u2010and\u2010neck intensity\u2010modulated radiation therapy (IMRT) plan generation without time\u2010consuming dose\u2010volume\u2010based inverse planning.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>This AI agent was trained via implementing a conditional generative adversarial network (cGAN) architecture. The generator, PyraNet, is a novel deep learning network that implements 28 classic ResNet blocks in pyramid\u2010like concatenations. The discriminator is a customized four\u2010layer DenseNet. The AI agent first generates multiple customized two\u2010dimensional projections at nine template beam angles from a patient\u2019s three\u2010dimensional computed tomography (CT) volume and structures. These projections are then stacked as four\u2010dimensional inputs of PyraNet, from which nine radiation fluence maps of the corresponding template beam angles are generated simultaneously. Finally, the predicted fluence maps are automatically postprocessed by Gaussian deconvolution operations and imported into a commercial treatment planning system (TPS) for plan integrity check and visualization. The AI agent was built and tested upon 231 oropharyngeal IMRT plans from a TPS plan library. 200\/16\/15 plans were assigned for training\/validation\/testing, respectively. Only the primary plans in the sequential boost regime were studied. All plans were normalized to 44\u00a0Gy prescription (2\u00a0Gy\/fx). A customized Harr wavelet loss was adopted for fluence map comparison during the training of the PyraNet. For test cases, isodose distributions in AI plans and TPS plans were qualitatively evaluated for overall dose distributions. Key dosimetric metrics were compared by Wilcoxon signed\u2010rank tests with a significance level of 0.05.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>\n                      All 15 AI plans were successfully generated. Isodose gradients outside of PTV in AI plans were comparable to those of the TPS plans. After PTV coverage normalization, D\n                      <jats:sub>mean<\/jats:sub>\n                      of left parotid (D\n                      <jats:sub>AI<\/jats:sub>\n                      \u00a0=\u00a023.1\u00a0\u00b1\u00a02.4\u00a0Gy; D\n                      <jats:sub>TPS<\/jats:sub>\n                      \u00a0=\u00a023.1\u00a0\u00b1\u00a02.0\u00a0Gy), right parotid (D\n                      <jats:sub>AI<\/jats:sub>\n                      \u00a0=\u00a023.8\u00a0\u00b1\u00a03.0\u00a0Gy; D\n                      <jats:sub>TPS<\/jats:sub>\n                      \u00a0=\u00a023.9\u00a0\u00b1\u00a02.3\u00a0Gy), and oral cavity (D\n                      <jats:sub>AI<\/jats:sub>\n                      \u00a0=\u00a024.7\u00a0\u00b1\u00a06.0\u00a0Gy; D\n                      <jats:sub>TPS<\/jats:sub>\n                      \u00a0=\u00a023.9\u00a0\u00b1\u00a04.3\u00a0Gy) in the AI plans and the TPS plans were comparable without statistical significance. AI plans achieved comparable results for maximum dose at 0.01cc of brainstem (D\n                      <jats:sub>AI<\/jats:sub>\n                      \u00a0=\u00a015.0\u00a0\u00b1\u00a02.1\u00a0Gy; D\n                      <jats:sub>TPS<\/jats:sub>\n                      \u00a0=\u00a015.5\u00a0\u00b1\u00a02.7\u00a0Gy) and cord\u00a0+\u00a05mm (D\n                      <jats:sub>AI<\/jats:sub>\n                      \u00a0=\u00a027.5\u00a0\u00b1\u00a02.3\u00a0Gy; D\n                      <jats:sub>TPS<\/jats:sub>\n                      \u00a0=\u00a025.8\u00a0\u00b1\u00a01.9\u00a0Gy) without clinically relevant differences, but body D\n                      <jats:sub>max<\/jats:sub>\n                      results (D\n                      <jats:sub>AI<\/jats:sub>\n                      \u00a0=\u00a0121.1\u00a0\u00b1\u00a03.9\u00a0Gy; D\n                      <jats:sub>TPS<\/jats:sub>\n                      \u00a0=\u00a0109.0\u00a0\u00b1\u00a00.9\u00a0Gy) were higher than the TPS plan results. The AI agent needed ~3\u00a0s for predicting fluence maps of an IMRT plan.\n                    <\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusions<\/jats:title>\n                    <jats:p>With rapid and fully automated execution, the developed AI agent can generate complex head\u2010and\u2010neck IMRT plans with acceptable dosimetry quality. This approach holds great potential for clinical applications in preplanning decision\u2010making and real\u2010time planning.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.1002\/mp.14770","type":"journal-article","created":{"date-parts":[[2021,2,12]],"date-time":"2021-02-12T12:58:48Z","timestamp":1613134728000},"page":"2714-2723","update-policy":"https:\/\/doi.org\/10.1002\/crossmark_policy","source":"Crossref","is-referenced-by-count":39,"title":["An artificial intelligence\u2010driven agent for real\u2010time head\u2010and\u2010neck IMRT plan generation using conditional generative adversarial network (cGAN)"],"prefix":"10.1002","volume":"48","author":[{"given":"Xinyi","family":"Li","sequence":"first","affiliation":[{"name":"Duke University Medical Center Durham NC 27710 USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Chunhao","family":"Wang","sequence":"additional","affiliation":[{"name":"Duke University Medical Center Durham NC 27710 USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yang","family":"Sheng","sequence":"additional","affiliation":[{"name":"Duke University Medical Center Durham NC 27710 USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jiahan","family":"Zhang","sequence":"additional","affiliation":[{"name":"Duke University Medical Center Durham NC 27710 USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Wentao","family":"Wang","sequence":"additional","affiliation":[{"name":"Duke University Medical Center Durham NC 27710 USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Fang\u2010Fang","family":"Yin","sequence":"additional","affiliation":[{"name":"Duke University Medical Center Durham NC 27710 USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Qiuwen","family":"Wu","sequence":"additional","affiliation":[{"name":"Duke University Medical Center Durham NC 27710 USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Q. Jackie","family":"Wu","sequence":"additional","affiliation":[{"name":"Duke University Medical Center Durham NC 27710 USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yaorong","family":"Ge","sequence":"additional","affiliation":[{"name":"University of North Carolina at Charlotte Charlotte NC 28223 USA"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"311","published-online":{"date-parts":[[2021,4,25]]},"reference":[{"key":"e_1_2_9_2_1","first-page":"7","volume-title":"Cancer Statistics, 2020","author":"Siegel RL","year":"2020"},{"key":"e_1_2_9_3_1","doi-asserted-by":"publisher","DOI":"10.6004\/jnccn.2017.0101"},{"key":"e_1_2_9_4_1","unstructured":"Network NCC.NCCN clinical practice guidelines in oncology: Head and neck cancers version.2013;2:2013."},{"key":"e_1_2_9_5_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.radonc.2015.04.021"},{"key":"e_1_2_9_6_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.ijrobp.2011.09.015"},{"key":"e_1_2_9_7_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.ijrobp.2017.02.215"},{"key":"e_1_2_9_8_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.ctro.2016.08.001"},{"key":"e_1_2_9_9_1","doi-asserted-by":"publisher","DOI":"10.1186\/s13014-018-1113-z"},{"key":"e_1_2_9_10_1","doi-asserted-by":"publisher","DOI":"10.1118\/1.4757927"},{"key":"e_1_2_9_11_1","doi-asserted-by":"publisher","DOI":"10.1177\/1533033819894577"},{"key":"e_1_2_9_12_1","doi-asserted-by":"publisher","DOI":"10.1259\/bjr.20180270"},{"key":"e_1_2_9_13_1","doi-asserted-by":"publisher","DOI":"10.1002\/mp.13679"},{"key":"e_1_2_9_14_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.ijrobp.2019.12.034"},{"key":"e_1_2_9_15_1","doi-asserted-by":"publisher","DOI":"10.1088\/0031-9155\/48\/17\/308"},{"key":"e_1_2_9_16_1","doi-asserted-by":"publisher","DOI":"10.1088\/0031-9155\/53\/4\/011"},{"key":"e_1_2_9_17_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.oraloncology.2018.03.020"},{"key":"e_1_2_9_18_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.meddos.2019.04.003"},{"key":"e_1_2_9_19_1","doi-asserted-by":"publisher","DOI":"10.2196\/14401"},{"key":"e_1_2_9_20_1","doi-asserted-by":"publisher","DOI":"10.1002\/mp.13597"},{"key":"e_1_2_9_21_1","doi-asserted-by":"publisher","DOI":"10.1088\/1361-6560\/aaef74"},{"key":"e_1_2_9_22_1","unstructured":"MahmoodR BabierA McNivenA DiamantA ChanTC.Automated treatment planning in radiation therapy using generative adversarial networks. arXiv preprint arXiv:06489.2018."},{"key":"e_1_2_9_23_1","doi-asserted-by":"publisher","DOI":"10.1002\/mp.13271"},{"key":"e_1_2_9_24_1","doi-asserted-by":"publisher","DOI":"10.1088\/1361-6560\/aa71f8"},{"key":"e_1_2_9_25_1","doi-asserted-by":"publisher","DOI":"10.1002\/mp.13896"},{"key":"e_1_2_9_26_1","doi-asserted-by":"publisher","DOI":"10.1002\/mp.13262"},{"key":"e_1_2_9_27_1","doi-asserted-by":"publisher","DOI":"10.1038\/s41598-018-37186-2"},{"key":"e_1_2_9_28_1","doi-asserted-by":"publisher","DOI":"10.1088\/1361-6560\/aba5eb"},{"key":"e_1_2_9_29_1","doi-asserted-by":"crossref","unstructured":"IsolaP ZhuJ\u2010Y ZhouT EfrosAA.Image\u2010to\u2010image translation with conditional adversarial networks.2017;1125\u20131134.","DOI":"10.1109\/CVPR.2017.632"},{"key":"e_1_2_9_30_1","first-page":"79","article-title":"The Haar\u2010wavelet transform in digital image processing: its status and achievements","volume":"13","author":"Porwik P","year":"2004","journal-title":"Mach Graph Vis"},{"key":"e_1_2_9_31_1","first-page":"e94","article-title":"Automatic IMRT planning via static field fluence prediction (AIP\u2010SFFP): a deep learning method for real\u2010time prostate treatment planning","volume":"46","author":"Li X","year":"2019","journal-title":"Med Phys"},{"key":"e_1_2_9_32_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.ijrobp.2019.06.760"},{"key":"e_1_2_9_33_1","unstructured":"HeK ZhangX RenS SunJ.Deep Residual Learning for Image Recognition. arXiv e\u2010prints.2015:arXiv:1512.03385. Accessed December 01 2015.https:\/\/ui.adsabs.harvard.edu\/abs\/2015arXiv151203385H"},{"key":"e_1_2_9_34_1","unstructured":"HuangG LiuZ van derMaatenL WeinbergerKQ.Densely Connected Convolutional Networks. arXiv e\u2010prints.2016:arXiv:1608.06993. Accessed August 01 2016.https:\/\/ui.adsabs.harvard.edu\/abs\/2016arXiv160806993H"},{"key":"e_1_2_9_35_1","doi-asserted-by":"publisher","DOI":"10.1016\/0360-3016(93)90548-A"},{"key":"e_1_2_9_36_1","doi-asserted-by":"publisher","DOI":"10.1016\/S0360-3016(02)04617-5"},{"key":"e_1_2_9_37_1","unstructured":"MirzaM OsinderoS.Conditional generative adversarial nets. arXiv preprint.2014."},{"key":"e_1_2_9_38_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.ijrobp.2016.06.2308"},{"key":"e_1_2_9_39_1","doi-asserted-by":"publisher","DOI":"10.1002\/mp.12556"}],"container-title":["Medical Physics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/onlinelibrary.wiley.com\/doi\/pdf\/10.1002\/mp.14770","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/onlinelibrary.wiley.com\/doi\/full-xml\/10.1002\/mp.14770","content-type":"application\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/aapm.onlinelibrary.wiley.com\/doi\/pdf\/10.1002\/mp.14770","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2023,8,23]],"date-time":"2023-08-23T15:11:07Z","timestamp":1692803467000},"score":1,"resource":{"primary":{"URL":"https:\/\/aapm.onlinelibrary.wiley.com\/doi\/10.1002\/mp.14770"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2021,4,25]]},"references-count":38,"journal-issue":{"issue":"6","published-print":{"date-parts":[[2021,6]]}},"alternative-id":["10.1002\/mp.14770"],"URL":"https:\/\/doi.org\/10.1002\/mp.14770","archive":["Portico"],"relation":{},"ISSN":["0094-2405","2473-4209"],"issn-type":[{"value":"0094-2405","type":"print"},{"value":"2473-4209","type":"electronic"}],"subject":[],"published":{"date-parts":[[2021,4,25]]},"assertion":[{"value":"2020-07-13","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2021-02-04","order":1,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2021-04-25","order":2,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}