000307471 001__ 307471
000307471 005__ 20251231120305.0
000307471 0247_ $$2doi$$a10.1088/1361-6560/ae2db9
000307471 0247_ $$2pmid$$apmid:41401509
000307471 0247_ $$2ISSN$$a0031-9155
000307471 0247_ $$2ISSN$$a1361-6560
000307471 037__ $$aDKFZ-2025-03070
000307471 041__ $$aEnglish
000307471 082__ $$a530
000307471 1001_ $$00009-0002-2650-4680$$aJulius Blöcker, Tom$$b0
000307471 245__ $$aGTV segmentation in MRI guided radiotherapy with promptable foundation models.
000307471 260__ $$aBristol$$bIOP Publ.$$c2025
000307471 3367_ $$2DRIVER$$aarticle
000307471 3367_ $$2DataCite$$aOutput Types/Journal article
000307471 3367_ $$0PUB:(DE-HGF)16$$2PUB:(DE-HGF)$$aJournal Article$$bjournal$$mjournal$$s1767100433_1386406
000307471 3367_ $$2BibTeX$$aARTICLE
000307471 3367_ $$2ORCID$$aJOURNAL_ARTICLE
000307471 3367_ $$00$$2EndNote$$aJournal Article
000307471 500__ $$aPUBLISHED29 December 2025
000307471 520__ $$aObjective. Magnetic resonance imaging (MRI) guided radiotherapy requires the delineation of gross tumor volumes (GTV) in daily MRI from MRI-linacs. Specialized models have been developed for this task for certain tumors. This study investigated an alternative, using promptable foundation models.Approach. Promptable foundation models were prompted with six different sparse geometric prompt types (points, boxes, 2D masks) to produce GTV segmentation masks, including Segment-anything 2 (SAM2), SAM2 fine-tuned for medical imaging (MedSAM2), and nnInteractive, an nnUnet-based promptable model for medical imaging. A diverse multi-institutional dataset of clinical GTV masks from the abdomen, lung, liver, pancreas, and pelvis sites on MRI scans from MRI-linacs was used to evaluate model outputs using various metrics, including the Dice similarity coefficient (DSC).Main results. The models produced segmentation masks comparable or superior to those from domain-specific models with median DSCs of up to 0.85 (nnInteractive-mask3 prompt). Prompts with more spatial information yielded better results with lower variance, with the effect reduced for nnInteractive and MedSAM2. These produced overall better results (median DSC over all prompt types 0.75 for nnInteractive, 0.70 for MedSAM2, 0.54 for SAM2).Significance. This investigation showed that promptable foundation models can in principle be used for GTV segmentation in MRI across multiple tumor types, although more research is necessary to reduce the variance and improve model performance.
000307471 536__ $$0G:(DE-HGF)POF4-899$$a899 - ohne Topic (POF4-899)$$cPOF4-899$$fPOF IV$$x0
000307471 588__ $$aDataset connected to CrossRef, PubMed, , Journals: inrepo02.dkfz.de
000307471 650_7 $$2Other$$aGTV segmentation
000307471 650_7 $$2Other$$aMRI-linac
000307471 650_7 $$2Other$$aMRgRT
000307471 650_7 $$2Other$$adeep learning
000307471 650_7 $$2Other$$afoundation models
000307471 650_7 $$2Other$$apromptable models
000307471 650_2 $$2MeSH$$aMagnetic Resonance Imaging
000307471 650_2 $$2MeSH$$aHumans
000307471 650_2 $$2MeSH$$aRadiotherapy, Image-Guided: methods
000307471 650_2 $$2MeSH$$aTumor Burden
000307471 650_2 $$2MeSH$$aImage Processing, Computer-Assisted: methods
000307471 650_2 $$2MeSH$$aNeoplasms: radiotherapy
000307471 650_2 $$2MeSH$$aNeoplasms: diagnostic imaging
000307471 650_2 $$2MeSH$$aNeoplasms: pathology
000307471 7001_ $$aDelopoulos, Nikolaos$$b1
000307471 7001_ $$aPalacios, Miguel A$$b2
000307471 7001_ $$00000-0003-3139-3444$$aKlüter, Sebastian$$b3
000307471 7001_ $$aHörner-Rieber, Juliane$$b4
000307471 7001_ $$aRippke, Carolin$$b5
000307471 7001_ $$aPlacidi, Lorenzo$$b6
000307471 7001_ $$aBoldrini, Luca$$b7
000307471 7001_ $$aFrascino, Vincenzo$$b8
000307471 7001_ $$aAndratschke, Nicolaus$$b9
000307471 7001_ $$aBaumgartl, Michael$$b10
000307471 7001_ $$aDal Bello, Riccardo$$b11
000307471 7001_ $$aMarschner, Sebastian N$$b12
000307471 7001_ $$0P:(DE-HGF)0$$aBelka, Claus$$b13
000307471 7001_ $$aCorradini, Stefanie$$b14
000307471 7001_ $$00000-0003-0667-3727$$aDudas, Denis$$b15
000307471 7001_ $$00000-0002-2431-4966$$aRiboldi, Marco$$b16
000307471 7001_ $$aKurz, Christopher$$b17
000307471 7001_ $$00000-0003-1707-4068$$aLandry, Guillaume$$b18
000307471 773__ $$0PERI:(DE-600)1473501-5$$a10.1088/1361-6560/ae2db9$$gVol. 71, no. 1, p. 015006 -$$n1$$p015006$$tPhysics in medicine and biology$$v71$$x0031-9155$$y2025
000307471 909CO $$ooai:inrepo02.dkfz.de:307471$$pVDB
000307471 9101_ $$0I:(DE-588b)2036810-0$$6P:(DE-HGF)0$$aDeutsches Krebsforschungszentrum$$b13$$kDKFZ
000307471 9131_ $$0G:(DE-HGF)POF4-899$$1G:(DE-HGF)POF4-890$$2G:(DE-HGF)POF4-800$$3G:(DE-HGF)POF4$$4G:(DE-HGF)POF$$aDE-HGF$$bProgrammungebundene Forschung$$lohne Programm$$vohne Topic$$x0
000307471 9141_ $$y2025
000307471 915__ $$0StatID:(DE-HGF)0430$$2StatID$$aNational-Konsortium$$d2024-12-27$$wger
000307471 915__ $$0StatID:(DE-HGF)0200$$2StatID$$aDBCoverage$$bSCOPUS$$d2024-12-27
000307471 915__ $$0StatID:(DE-HGF)0300$$2StatID$$aDBCoverage$$bMedline$$d2024-12-27
000307471 915__ $$0StatID:(DE-HGF)0199$$2StatID$$aDBCoverage$$bClarivate Analytics Master Journal List$$d2024-12-27
000307471 915__ $$0StatID:(DE-HGF)1050$$2StatID$$aDBCoverage$$bBIOSIS Previews$$d2024-12-27
000307471 915__ $$0StatID:(DE-HGF)0160$$2StatID$$aDBCoverage$$bEssential Science Indicators$$d2024-12-27
000307471 915__ $$0StatID:(DE-HGF)1030$$2StatID$$aDBCoverage$$bCurrent Contents - Life Sciences$$d2024-12-27
000307471 915__ $$0StatID:(DE-HGF)1190$$2StatID$$aDBCoverage$$bBiological Abstracts$$d2024-12-27
000307471 915__ $$0StatID:(DE-HGF)0113$$2StatID$$aWoS$$bScience Citation Index Expanded$$d2024-12-27
000307471 915__ $$0StatID:(DE-HGF)0150$$2StatID$$aDBCoverage$$bWeb of Science Core Collection$$d2024-12-27
000307471 915__ $$0StatID:(DE-HGF)0100$$2StatID$$aJCR$$bPHYS MED BIOL : 2022$$d2024-12-27
000307471 915__ $$0StatID:(DE-HGF)0600$$2StatID$$aDBCoverage$$bEbsco Academic Search$$d2024-12-27
000307471 915__ $$0StatID:(DE-HGF)0030$$2StatID$$aPeer Review$$bASC$$d2024-12-27
000307471 915__ $$0StatID:(DE-HGF)9900$$2StatID$$aIF < 5$$d2024-12-27
000307471 9201_ $$0I:(DE-He78)MU01-20160331$$kMU01$$lDKTK Koordinierungsstelle München$$x0
000307471 980__ $$ajournal
000307471 980__ $$aVDB
000307471 980__ $$aI:(DE-He78)MU01-20160331
000307471 980__ $$aUNRESTRICTED