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

László Ruskó, Marta Capala, Vanda Czipczer, Bernadett Kolozsvári, Borbála Deák-Karancsi, Renáta Czabány, Bence Gyalai, Tao Tan, Zoltán Végváry, Emőke Borzasi, Zsófia Együd, Renáta Kószó, Viktor Paczona, Emese Fodor, Chad Bobb, Cristina Cozzini, Sandeep Kaushik, Barbara Darázs, Gerda Verduijn, Rachel Pearson, Ross Maxwell, Hazel Mccallum, Juan Hernandez Tamames, Katalin Hideghéty, Steven Petit, Florian Wiesinger

Abstract

Segmentation of organs-at-risk (OAR) in MR images has several clinical applications; including radiation therapy (RT) planning. This paper presents a deep-learning-based method to segment 15 structures in the head region. The proposed method first applies 2D U-Net models to each of the three planes (axial, coronal, sagittal) to roughly segment the structure. Then, the results of the 2D models are combined into a fused prediction to localize the 3D bounding box of the structure. Finally, a 3D U-Net is applied to the volume of the bounding box to determine the precise contour of the structure. The model was trained on a public dataset and evaluated on both public and private datasets that contain T2-weighted MR scans of the head-and-neck region. For all cases the contour of each structure was defined by operators trained by expert clinical delineators. The evaluation demonstrated that various structures can be accurately and efficiently localized and segmented using the presented framework. The contours generated by the proposed method were also qualitatively evaluated. The majority (92%) of the segmented OARs was rated as clinically useful for radiation therapy.

Download


Paper Citation


in Harvard Style

Ruskó L., Capala M., Czipczer V., Kolozsvári B., Deák-Karancsi B., Czabány R., Gyalai B., Tan T., Végváry Z., Borzasi E., Együd Z., Kószó R., Paczona V., Fodor E., Bobb C., Cozzini C., Kaushik S., Darázs B., Verduijn G., Pearson R., Maxwell R., Mccallum H., Hernandez Tamames J., Hideghéty K., Petit S. and Wiesinger F. (2021). Deep-Learning-based Segmentation of Organs-at-Risk in the Head for MR-assisted Radiation Therapy Planning.In Proceedings of the 14th International Joint Conference on Biomedical Engineering Systems and Technologies - Volume 1: BIOIMAGING, ISBN 978-989-758-490-9, pages 31-43. DOI: 10.5220/0010235000310043


in Bibtex Style

@conference{bioimaging21,
author={László Ruskó and Marta Capala and Vanda Czipczer and Bernadett Kolozsvári and Borbála Deák-Karancsi and Renáta Czabány and Bence Gyalai and Tao Tan and Zoltán Végváry and Emőke Borzasi and Zsófia Együd and Renáta Kószó and Viktor Paczona and Emese Fodor and Chad Bobb and Cristina Cozzini and Sandeep Kaushik and Barbara Darázs and Gerda Verduijn and Rachel Pearson and Ross Maxwell and Hazel Mccallum and Juan Hernandez Tamames and Katalin Hideghéty and Steven Petit and Florian Wiesinger},
title={Deep-Learning-based Segmentation of Organs-at-Risk in the Head for MR-assisted Radiation Therapy Planning},
booktitle={Proceedings of the 14th International Joint Conference on Biomedical Engineering Systems and Technologies - Volume 1: BIOIMAGING,},
year={2021},
pages={31-43},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0010235000310043},
isbn={978-989-758-490-9},
}


in EndNote Style

TY - CONF

JO - Proceedings of the 14th International Joint Conference on Biomedical Engineering Systems and Technologies - Volume 1: BIOIMAGING,
TI - Deep-Learning-based Segmentation of Organs-at-Risk in the Head for MR-assisted Radiation Therapy Planning
SN - 978-989-758-490-9
AU - Ruskó L.
AU - Capala M.
AU - Czipczer V.
AU - Kolozsvári B.
AU - Deák-Karancsi B.
AU - Czabány R.
AU - Gyalai B.
AU - Tan T.
AU - Végváry Z.
AU - Borzasi E.
AU - Együd Z.
AU - Kószó R.
AU - Paczona V.
AU - Fodor E.
AU - Bobb C.
AU - Cozzini C.
AU - Kaushik S.
AU - Darázs B.
AU - Verduijn G.
AU - Pearson R.
AU - Maxwell R.
AU - Mccallum H.
AU - Hernandez Tamames J.
AU - Hideghéty K.
AU - Petit S.
AU - Wiesinger F.
PY - 2021
SP - 31
EP - 43
DO - 10.5220/0010235000310043