{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,11,6]],"date-time":"2026-11-06T11:46:46Z","timestamp":1793965606691,"version":"4.8.0"},"reference-count":43,"publisher":"IOP Publishing","issue":"15","license":[{"start":{"date-parts":[[2025,7,24]],"date-time":"2025-07-24T00:00:00Z","timestamp":1753315200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/iopscience.iop.org\/page\/copyright"},{"start":{"date-parts":[[2025,7,24]],"date-time":"2025-07-24T00:00:00Z","timestamp":1753315200000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/iopscience.iop.org\/info\/page\/text-and-data-mining"}],"content-domain":{"domain":["iopscience.iop.org"],"crossmark-restriction":false},"short-container-title":["Phys. Med. Biol."],"published-print":{"date-parts":[[2025,8,3]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>\n                    <jats:italic>Objective<\/jats:italic>\n                    . Radiotherapy treatment planning is a time-consuming and potentially subjective process that requires iterative adjustment of model parameters to balance multiple conflicting objectives. Recent advancements in frontier artificial intelligence (AI) models offer promising avenues for addressing the challenges in planning and clinical decision-making. This study introduces GPT-RadPlan, an automated treatment planning framework that integrates radiation oncology knowledge with the reasoning capabilities of large multi-modal models, such as GPT-4Vision (GPT-4 V) from OpenAI.\n                    <jats:italic>Approach<\/jats:italic>\n                    . Via in-context learning, we incorporate clinical requirements and three approved clinical plans along with their optimization settings, enabling GPT-4 V to acquire treatment planning domain knowledge. The resulting GPT-RadPlan system is integrated into our in-house inverse treatment planning system through an application programming interface. For a given patient, GPT-RadPlan acts as both plan evaluator and planner, first assessing dose distributions and dose-volume histograms, and then providing \u2018textual feedback\u2019 on how to improve the plan to match the physician\u2019s requirements. In this manner, GPT-RadPlan iteratively refines the plan by adjusting planning parameters, such as weights and dose objectives, based on its suggestions.\n                    <jats:italic>Main results<\/jats:italic>\n                    . The efficacy of the automated planning system is showcased across 17 prostate cancer and 13 head &amp; neck cancer VMAT plans with prescribed doses of 70.2\u2009Gy and 72\u2009Gy, respectively, where we compared GPT-RadPlan results to clinical plans produced by human experts. In all cases, GPT-RadPlan either outperformed or matched the clinical plans, demonstrating superior target coverage and reducing organ-at-risk doses by 5\u2009Gy on average (15% for prostate and 10%\u201315% for head &amp; neck).\n                    <jats:italic>Significance<\/jats:italic>\n                    . Consistently satisfying the the dose-volume objectives in the clinical protocol, GPT-RadPlan represents the first multimodal large language model agent that mimics the behaviors of human planners in radiation oncology clinics, achieving promising results in automating the treatment planning process without the need for additional training.\n                  <\/jats:p>","DOI":"10.1088\/1361-6560\/adf02c","type":"journal-article","created":{"date-parts":[[2025,7,15]],"date-time":"2025-07-15T22:55:27Z","timestamp":1752620127000},"page":"155002","update-policy":"https:\/\/doi.org\/10.1088\/crossmark-policy","source":"Crossref","is-referenced-by-count":21,"title":["Automated radiotherapy treatment planning guided by GPT-4Vision"],"prefix":"10.1088","volume":"70","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-6291-0550","authenticated-orcid":true,"given":"Sheng","family":"Liu","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-2328-1429","authenticated-orcid":true,"given":"Oscar","family":"Pastor-Serrano","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8799-2580","authenticated-orcid":true,"given":"Yizheng","family":"Chen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0004-1645-5912","authenticated-orcid":true,"given":"Matthew","family":"Gopaulchan","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-9924-693X","authenticated-orcid":true,"given":"Weixing","family":"Liang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Mark","family":"Buyyounouski","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Erqi","family":"Pollom","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Quynh-Thu","family":"Le","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4897-3843","authenticated-orcid":true,"given":"Michael","family":"Gensheimer","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Peng","family":"Dong","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yong","family":"Yang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"James","family":"Zou","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-2536-5359","authenticated-orcid":true,"given":"Lei","family":"Xing","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"266","published-online":{"date-parts":[[2025,7,24]]},"reference":[{"key":"pmbadf02cbib1","first-page":"pp 2623","article-title":"Optuna: a next-generation hyperparameter optimization framework","author":"Akiba","year":"2019"},{"key":"pmbadf02cbib2","doi-asserted-by":"publisher","first-page":"6339","DOI":"10.1088\/0031-9155\/52\/20\/016","article-title":"A novel approach to multi-criteria inverse planning for IMRT","volume":"52","author":"Breedveld","year":"2007","journal-title":"Phys. Med. Biol."},{"key":"pmbadf02cbib3","doi-asserted-by":"publisher","first-page":"951","DOI":"10.1118\/1.3676689","article-title":"iCycle: integrated, multicriterial beam angle and profile optimization for generation of coplanar and noncoplanar IMRT plans","volume":"39","author":"Breedveld","year":"2012","journal-title":"Med. Phys."},{"key":"pmbadf02cbib4","doi-asserted-by":"publisher","first-page":"21NT01","DOI":"10.1088\/1361-6560\/aae65e","article-title":"Continuous aperture dose calculation and optimization for volumetric modulated arc therapy","volume":"63","author":"Christiansen","year":"2018","journal-title":"Phys. Med. Biol."},{"key":"pmbadf02cbib5","doi-asserted-by":"publisher","first-page":"370","DOI":"10.1002\/mp.13271","article-title":"Automatic treatment planning based on three-dimensional dose distribution predicted from deep learning technique","volume":"46","author":"Fan","year":"2019","journal-title":"Med. Phys."},{"key":"pmbadf02cbib6","article-title":"Generative Adversarial Nets","volume":"vol 27","author":"Goodfellow","year":"2014"},{"key":"pmbadf02cbib7","doi-asserted-by":"publisher","first-page":"6140","DOI":"10.1002\/mp.14544","article-title":"Artificial intelligence-based radiotherapy machine parameter optimization using reinforcement learning","volume":"47","author":"Hrinivich","year":"2020","journal-title":"Med. Phys."},{"key":"pmbadf02cbib8","doi-asserted-by":"publisher","DOI":"10.1088\/1361-6560\/ac5672","article-title":"Meta-optimization for fully automated radiation therapy treatment planning","volume":"67","author":"Huang","year":"2022","journal-title":"Phys. Med. Biol."},{"key":"pmbadf02cbib9","doi-asserted-by":"publisher","DOI":"10.1088\/1361-6560\/accdb0","article-title":"Learning image representations for content-based image retrieval of radiotherapy treatment plans","volume":"68","author":"Huang","year":"2023","journal-title":"Phys. Med. Biol."},{"key":"pmbadf02cbib10","doi-asserted-by":"publisher","DOI":"10.1259\/bjr.20180270","article-title":"Automation in intensity modulated radiotherapy treatment planning\u2014a review of recent innovations","volume":"91","author":"Hussein","year":"2018","journal-title":"Br. J. Radiol."},{"key":"pmbadf02cbib11","first-page":"pp 13746","article-title":"Deep probability estimation","author":"Liu","year":"2022a"},{"key":"pmbadf02cbib12","article-title":"In-context vectors: making in context learning more effective and controllable through latent space steering","author":"Liu","year":"2024"},{"key":"pmbadf02cbib13","article-title":"Avoiding spurious correlations via logit correction","author":"Liu","year":"2022"},{"key":"pmbadf02cbib14","first-page":"pp 2638","article-title":"Calibrating llm-based evaluator","author":"Liu","year":"2024"},{"key":"pmbadf02cbib15","doi-asserted-by":"publisher","DOI":"10.1002\/mp.17985)","article-title":"RadOnc-GPT: a large language model for radiation oncology","author":"Liu","year":"2023"},{"key":"pmbadf02cbib16","doi-asserted-by":"publisher","first-page":"2978","DOI":"10.1002\/mp.13618","article-title":"Dose distribution prediction in isodose feature-preserving voxelization domain using deep convolutional neural network","volume":"46","author":"Ma","year":"2019","journal-title":"Med. Phys."},{"key":"pmbadf02cbib17","doi-asserted-by":"publisher","DOI":"10.1088\/1361-6560\/ab2146","article-title":"Incorporating dosimetric features into the prediction of 3D VMAT dose distributions using deep convolutional neural network","volume":"64","author":"Ma","year":"2019","journal-title":"Phys. Med. Biol."},{"key":"pmbadf02cbib18","doi-asserted-by":"publisher","first-page":"1558","DOI":"10.1002\/mp.14074","article-title":"Technical note: a fast inverse direct aperture optimization algorithm for volumetric-modulated arc therapy","volume":"47","author":"MacFarlane","year":"2020","journal-title":"Med. Phys."},{"key":"pmbadf02cbib19","doi-asserted-by":"publisher","first-page":"3679","DOI":"10.1002\/mp.13597","article-title":"Three-dimensional dose prediction for lung IMRT patients with deep neural networks: Robust learning from heterogeneous beam configurations","volume":"46","author":"Maria Barrag\u00e1n-Montero","year":"2019","journal-title":"Med. Phys."},{"key":"pmbadf02cbib20","doi-asserted-by":"publisher","first-page":"1000","DOI":"10.1109\/TMI.2015.2505188","article-title":"Contextual atlas regression forests: multiple-atlas-based automated dose prediction in radiation therapy","volume":"35","author":"McIntosh","year":"2016","journal-title":"IEEE Trans. Med. Imaging"},{"key":"pmbadf02cbib21","doi-asserted-by":"publisher","first-page":"6923","DOI":"10.1088\/0031-9155\/60\/17\/6923","article-title":"Dose-volume histogram prediction using density estimation","volume":"60","author":"Munter","year":"2015","journal-title":"Phys. Med. Biol."},{"key":"pmbadf02cbib22","doi-asserted-by":"publisher","first-page":"1076","DOI":"10.1038\/s41598-018-37741-x","article-title":"A feasibility study for predicting optimal radiation therapy dose distributions of prostate cancer patients from patient anatomy using deep learning","volume":"9","author":"Nguyen","year":"2019","journal-title":"Sci. Rep."},{"key":"pmbadf02cbib23","doi-asserted-by":"publisher","first-page":"113","DOI":"10.1016\/S0958-3947(01)00057-7","article-title":"Inverse planning for photon and proton beams","volume":"26","author":"Oelfke","year":"2001","journal-title":"Med. Dosimetry"},{"key":"pmbadf02cbib24","doi-asserted-by":"publisher","DOI":"10.1088\/1361-6560\/acc71d","article-title":"A probabilistic deep learning model of inter-fraction anatomical variations in radiotherapy","volume":"68","author":"Pastor-Serrano","year":"2023","journal-title":"Phys. Med. Biol."},{"key":"pmbadf02cbib25","doi-asserted-by":"publisher","DOI":"10.1088\/1361-6560\/ac383f","article-title":"How should we model and evaluate breathing interplay effects in IMPT?","volume":"66","author":"Pastor-Serrano","year":"2021","journal-title":"Phys. Med. Biol."},{"key":"pmbadf02cbib26","doi-asserted-by":"publisher","first-page":"17","DOI":"10.1016\/j.artmed.2016.01.006","article-title":"Knowledge-light adaptation approaches in case-based reasoning for radiotherapy treatment planning","volume":"68","author":"Petrovic","year":"2016","journal-title":"Artif. Intell. Med."},{"key":"pmbadf02cbib27","first-page":"pp 234","article-title":"U-Net: convolutional networks for biomedical image segmentation","author":"Ronneberger","year":"2015"},{"key":"pmbadf02cbib28","first-page":"pp 33","article-title":"An ontology-based adaptive personalized e-learning system, assisted by software agents on cloud storage. Knowledge-Based Systems","volume":"vol 90","author":"Russel","year":"2002"},{"key":"pmbadf02cbib29","doi-asserted-by":"publisher","first-page":"2329","DOI":"10.1002\/mp.14114","article-title":"Operating a treatment planning system using a deep-reinforcement learning-based virtual treatment planner for prostate cancer intensity-modulated radiation therapy treatment planning","volume":"47","author":"Shen","year":"2020","journal-title":"Med. Phys."},{"key":"pmbadf02cbib30","doi-asserted-by":"publisher","first-page":"4858","DOI":"10.1118\/1.4736803","article-title":"Integrated multicriterial optimization of beam angles and intensity profiles for coplanar and noncoplanar head and neck IMRT and implications for VMAT","volume":"39","author":"Voet","year":"2012","journal-title":"Med. Phys."},{"key":"pmbadf02cbib31","doi-asserted-by":"publisher","first-page":"866","DOI":"10.1016\/j.ijrobp.2012.04.015","article-title":"Toward fully automated multicriterial plan generation: a prospective clinical study","volume":"85","author":"Voet","year":"2013","journal-title":"Int. J. Radiat. Oncol. Biol. Phys."},{"key":"pmbadf02cbib32","doi-asserted-by":"publisher","first-page":"3773","DOI":"10.1002\/mp.16289","article-title":"High-dimensional automated radiation therapy treatment planning via Bayesian optimization","volume":"50","author":"Wang","year":"2023","journal-title":"Med. Phys."},{"key":"pmbadf02cbib33","doi-asserted-by":"publisher","DOI":"10.1088\/1361-6560\/adbff1","article-title":"A feasibility study of automating radiotherapy planning with large language model agents","volume":"70","author":"Wang","year":"2025","journal-title":"Phys. Med. Biol."},{"key":"pmbadf02cbib34","doi-asserted-by":"publisher","first-page":"678","DOI":"10.1259\/bjr\/65676879","article-title":"The physical basis of IMRT and inverse planning","volume":"76","author":"Webb","year":"2003","journal-title":"Br. J. Radiol."},{"key":"pmbadf02cbib35","first-page":"p 729","volume":"vol 12","author":"Wiering","year":"2012"},{"key":"pmbadf02cbib36","doi-asserted-by":"publisher","first-page":"2556","DOI":"10.1002\/mp.12251","article-title":"Development of the open-source dose calculation and optimization toolkit matRad","volume":"44","author":"Wieser","year":"2017","journal-title":"Med. Phys."},{"key":"pmbadf02cbib37","doi-asserted-by":"publisher","first-page":"2107","DOI":"10.1088\/0031-9155\/41\/10\/018","article-title":"Iterative methods for inverse treatment planning","volume":"41","author":"Xing","year":"1996","journal-title":"Phys. Med. Biol."},{"key":"pmbadf02cbib38","doi-asserted-by":"publisher","first-page":"2525","DOI":"10.1088\/0031-9155\/44\/10\/311","article-title":"Optimization of importance factors in inverse planning","volume":"44","author":"Xing","year":"1999","journal-title":"Phys. Med. Biol."},{"key":"pmbadf02cbib39","doi-asserted-by":"publisher","first-page":"22TR02","DOI":"10.1088\/1361-6560\/ac9cb3","article-title":"Deep reinforcement learning and its applications in medical imaging and radiation therapy: a survey","volume":"67","author":"Xu","year":"2022","journal-title":"Phys. Med. Biol."},{"key":"pmbadf02cbib40","doi-asserted-by":"publisher","first-page":"e364","DOI":"10.1016\/j.ijrobp.2020.07.2365","article-title":"Evaluation of a knowledge-guided automated treatment planning tool","volume":"108","author":"Yang","year":"2020","journal-title":"Int. J. Radiat. Oncol. Biol. Phys."},{"key":"pmbadf02cbib41","doi-asserted-by":"publisher","first-page":"5101","DOI":"10.1088\/0031-9155\/49\/22\/006","article-title":"Clinical knowledge-based inverse treatment planning","volume":"49","author":"Yang","year":"2004","journal-title":"Phys. Med. Biol."},{"key":"pmbadf02cbib42","doi-asserted-by":"publisher","first-page":"R31","DOI":"10.1088\/0031-9155\/56\/5\/R01","article-title":"Intensity-modulated arc therapy: principles, technologies and clinical implementation","volume":"56","author":"Yu","year":"2011","journal-title":"Phys. Med. Biol."},{"key":"pmbadf02cbib43","doi-asserted-by":"publisher","first-page":"543","DOI":"10.1007\/s10462-016-9491-9","article-title":"Learning from crowdsourced labeled data: a survey","volume":"46","author":"Zhang","year":"2016","journal-title":"Artif. Intell. Rev."}],"container-title":["Physics in Medicine &amp; Biology"],"original-title":[],"link":[{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6560\/adf02c","content-type":"text\/html","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6560\/adf02c\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6560\/adf02c","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6560\/adf02c\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6560\/adf02c\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6560\/adf02c\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6560\/adf02c\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"similarity-checking"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6560\/adf02c\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,7,24]],"date-time":"2025-07-24T09:15:04Z","timestamp":1753348504000},"score":1,"resource":{"primary":{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6560\/adf02c"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,7,24]]},"references-count":43,"journal-issue":{"issue":"15","published-online":{"date-parts":[[2025,7,24]]},"published-print":{"date-parts":[[2025,8,3]]}},"URL":"https:\/\/doi.org\/10.1088\/1361-6560\/adf02c","relation":{},"ISSN":["0031-9155","1361-6560"],"issn-type":[{"value":"0031-9155","type":"print"},{"value":"1361-6560","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,7,24]]},"assertion":[{"value":"Automated radiotherapy treatment planning guided by GPT-4Vision","name":"article_title","label":"Article Title"},{"value":"Physics in Medicine & Biology","name":"journal_title","label":"Journal Title"},{"value":"paper","name":"article_type","label":"Article Type"},{"value":"\u00a9 2025 Institute of Physics and Engineering in Medicine. All rights, including for text and data mining, AI training, and similar technologies, are reserved.","name":"copyright_information","label":"Copyright Information"},{"value":"2025-02-20","name":"date_received","label":"Date Received","group":{"name":"publication_dates","label":"Publication dates"}},{"value":"2025-07-15","name":"date_accepted","label":"Date Accepted","group":{"name":"publication_dates","label":"Publication dates"}},{"value":"2025-07-24","name":"date_epub","label":"Online publication date","group":{"name":"publication_dates","label":"Publication dates"}}]}}