Deep-Learning-based Segmentation of Organs-at-Risk in the Head for MR-assisted Radiation Therapy Planning

Ruskó, László; Capala, Marta E.; Czipczer, Vanda; Kolozsvári, Bernadett; Deák-Karancsi, Borbála; Czabány, Renáta; Gyalai, Bence; Tan, Tao; Végváry, Zoltán; Borzási, Emőke; Együd, Zsófia; Kószó, Renáta; Paczona, Viktor; Fodor, Emese; Bobb, Chad; Cozzini, C.; Kaushik, Sandeep; Darázs, Barbara; Verduijn, Gerda M.; Pearson, R.; Maxwell, Ross J.; McCallum, H.; Tamames, Juan Antonio Hernández; Hidéghety, Katalin; Петит, С.

doi:10.5220/0010235000310043

Cited by 5 publications

(5 citation statements)

References 1 publication

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A radiation oncologist working at the LMU MR‐Linac was presented two sets of contours in random order: the predicted and the propagated, for each fraction. First, the physician was asked to choose the contour considered more useful during plan adaptation, and secondly, to rate each delineation on a four‐point scale: 1‐ready to use, 2‐small corrections required, 3‐major corrections required, and 4‐not useful 33 . In order to eliminate personal bias, the physician was neither informed about the study goal nor the origin of the examined delineations.…”

Section: Methodsmentioning

confidence: 99%

“…First, the physician was asked to choose the contour considered more useful during plan adaptation, and secondly, to rate each delineation on a four-point scale: 1-ready to use, 2-small corrections required, 3-major corrections required, and 4-not useful. 33 In order to eliminate personal bias, the physician was neither informed about the study goal nor the origin of the examined delineations. Since CTV segmentation requires additional knowledge, such as the patient's medical record and cancer risk category, this analysis was restricted to the OARs.…”

Section: Network-predicted Versus Treatment Planning System-propagate...mentioning

confidence: 99%

See 1 more Smart Citation

Patient‐specific transfer learning for auto‐segmentation in adaptive 0.35 T MRgRT of prostate cancer: a bi‐centric evaluation

et al. 2022

View full text Add to dashboard Cite

Background: Online adaptive radiation therapy (RT) using hybrid magnetic resonance linear accelerators (MR-Linacs) can administer a tailored radiation dose at each treatment fraction. Daily MR imaging followed by organ and target segmentation adjustments allow to capture anatomical changes, improve target volume coverage, and reduce the risk of side effects. The introduction of automatic segmentation techniques could help to further improve the online adaptive workflow by shortening the re-contouring time and reducing intra-and inter-observer variability. In fractionated RT, prior knowledge, such as planning images and manual expert contours, is usually available before irradiation, but not used by current artificial intelligence-based autocontouring approaches. Purpose: The goal of this study was to train convolutional neural networks (CNNs) for automatic segmentation of bladder, rectum (organs at risk, OARs), and clinical target volume (CTV) for prostate cancer patients treated at 0.35 T MR-Linacs. Furthermore, we tested the CNNs generalization on data from independent facilities and compared them with the MR-Linac treatment planning system (TPS) propagated structures currently used in clinics. Finally, expert planning delineations were utilized for patient-(PS) and facility-specific (FS) transfer learning to improve auto-segmentation of CTV and OARs on fraction images. Methods: In this study, data from fractionated treatments at 0.35 T MR-Linacs were leveraged to develop a 3D U-Net-based automatic segmentation. Cohort C1 had 73 planning images and cohort C2 had 19 planning and 240 fraction images. The baseline models (BMs) were trained solely on C1 planning data using 53 MRIs for training and 10 for validation. To assess their accuracy, the models were tested on three data subsets: (i) 10 C1 planning images not used for training, (ii) 19 C2 planning, and (iii) 240 C2 fraction images. BMs also served as a starting point for FS and PS transfer learning, where the planning images from C2 were used for network parameter fine tuning. The segmentation output of the different trained models was compared against expert ground truth by means of geometric metrics. Moreover, a trained physician graded the network segmentations as well as the segmentations propagated by the clinical TPS.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Network-predicted Versus Treatment Planning System-propagate...mentioning

confidence: 99%

Patient‐specific transfer learning for auto‐segmentation in adaptive 0.35 T MRgRT of prostate cancer: a bi‐centric evaluation

et al. 2022

View full text Add to dashboard Cite

show abstract

“…10 Previous studies of autocontouring for brain OARs using deep learning relied only on geometric assessment. 7,[11][12][13] By evaluating the correlation between geometric and dosimetric measures, we aim to establish whether geometric assessment alone is sufficient to evaluate brain OAR autosegmentation tools or whether an additional dosimetric evaluation is also needed.…”

Section: Introductionmentioning

confidence: 99%

Dosimetric impact of contour editing on CT and MRI deep‐learning autosegmentation for brain OARs

Alzahrani,

Henry,

Clark

et al. 2024

J Applied Clin Med Phys

View full text Add to dashboard Cite

PurposeTo establish the clinical applicability of deep‐learning organ‐at‐risk autocontouring models (DL‐AC) for brain radiotherapy. The dosimetric impact of contour editing, prior to model training, on performance was evaluated for both CT and MRI‐based models. The correlation between geometric and dosimetric measures was also investigated to establish whether dosimetric assessment is required for clinical validation.MethodCT and MRI‐based deep learning autosegmentation models were trained using edited and unedited clinical contours. Autosegmentations were dosimetrically compared to gold standard contours for a test cohort. D1%, D5%, D50%, and maximum dose were used as clinically relevant dosimetric measures. The statistical significance of dosimetric differences between the gold standard and autocontours was established using paired Student's t‐tests. Clinically significant cases were identified via dosimetric headroom to the OAR tolerance. Pearson's Correlations were used to investigate the relationship between geometric measures and absolute percentage dose changes for each autosegmentation model.ResultsExcept for the right orbit, when delineated using MRI models, the dosimetric statistical analysis revealed no superior model in terms of the dosimetric accuracy between the CT DL‐AC models or between the MRI DL‐AC for any investigated brain OARs. The number of patients where the clinical significance threshold was exceeded was higher for the optic chiasm D1% than other OARs, for all autosegmentation models. A weak correlation was consistently observed between the outcomes of dosimetric and geometric evaluations.ConclusionsEditing contours before training the DL‐AC model had no significant impact on dosimetry. The geometric test metrics were inadequate to estimate the impact of contour inaccuracies on dose. Accordingly, dosimetric analysis is needed to evaluate the clinical applicability of DL‐AC models in the brain.

show abstract

“…One of the most popular CNN architectures is the U-Net [10] and its extension, the 3D U-Net [11]. They are widely used in several publications for automatic OARs segmentation and demonstrated valuable outcomes [12][13][14][15][16][17][18][19]. [12,13] both presented a segmentation framework that localizes and then segments 8 and 6 head-and-neck OARs, respectively.…”

Section: Introductionmentioning

confidence: 99%

“…[14] performs segmentation on 8 structures using an ensemble of multi-class 2D U-Nets and a graph-based postprocessing. In [15], a two-stage deep learning basedsegmentation algorithm is proposed for 8 head OARs which uses 2D U-Nets-based localization followed by a 3D U-Net model to finely segment the cropped smaller area. In [16], for the automatic delineation of submandibular glands, parotid glands and level II and level III lymph nodes, a 3D U-Net was used.…”

Section: Introductionmentioning

confidence: 99%

Comprehensive deep learning-based framework for automatic organs-at-risk segmentation in head-and-neck and pelvis for MR-guided radiation therapy planning

Czipczer,

Kolozsvári,

Deák-Karancsi

et al. 2023

Front. Phys.

View full text Add to dashboard Cite

Introduction: The excellent soft-tissue contrast of magnetic resonance imaging (MRI) is appealing for delineation of organs-at-risk (OARs) as it is required for radiation therapy planning (RTP). In the last decade there has been an increasing interest in using deep-learning (DL) techniques to shorten the labor-intensive manual work and increase reproducibility. This paper focuses on the automatic segmentation of 27 head-and-neck and 10 male pelvis OARs with deep-learning methods based on T2-weighted MR images.Method: The proposed method uses 2D U-Nets for localization and 3D U-Net for segmentation of the various structures. The models were trained using public and private datasets and evaluated on private datasets only.Results and discussion: Evaluation with ground-truth contours demonstrated that the proposed method can accurately segment the majority of OARs and indicated similar or superior performance to state-of-the-art models. Furthermore, the auto-contours were visually rated by clinicians using Likert score and on average, 81% of them was found clinically acceptable.

show abstract

Deep-Learning-based Segmentation of Organs-at-Risk in the Head for MR-assisted Radiation Therapy Planning

Cited by 5 publications

References 1 publication

Patient‐specific transfer learning for auto‐segmentation in adaptive 0.35 T MRgRT of prostate cancer: a bi‐centric evaluation

Patient‐specific transfer learning for auto‐segmentation in adaptive 0.35 T MRgRT of prostate cancer: a bi‐centric evaluation

Dosimetric impact of contour editing on CT and MRI deep‐learning autosegmentation for brain OARs

Comprehensive deep learning-based framework for automatic organs-at-risk segmentation in head-and-neck and pelvis for MR-guided radiation therapy planning

Contact Info

Product

Resources

About