Abstract
For the head and neck (HaN) cancer, radiotherapy is a mainstay treatment modality that aims to deliver a high radiation dose to the targeted cancerous cells while sparing the nearby healthy organs-at-risk (OARs). A precise three-dimensional segmentation of OARs from computed tomography (CT) images is required for optimal radiation dose distribution calculation, however, so far there has been no evaluation about the impact of the combined analysis of multiple imaging modalities, such as CT and magnetic resonance (MR). For this purpose, we have devised a database of 56 CT and MR images of the same patients with 31 manually delineated OARs, and in this paper we present the baseline segmentation results that were obtained by applying the nnU-Net framework. The resulting average Dice coefficient of 68% and average 95-percentile Hausdorff distance of 8.2mm on a subset of 14 images indicate that nnU-Net serves as a solid baseline method.
Original language | English |
---|---|
Title of host publication | 2022 IEEE 19th International Symposium on Biomedical Imaging (ISBI) |
Publisher | IEEE |
Publication date | 2022 |
Pages | 1-4 |
ISBN (Electronic) | 9781665429238 |
DOIs | |
Publication status | Published - 2022 |
Event | 19th IEEE International Symposium on Biomedical Imaging, ISBI 2022 - Kolkata, India Duration: 28 Mar 2022 → 31 Mar 2022 |
Conference
Conference | 19th IEEE International Symposium on Biomedical Imaging, ISBI 2022 |
---|---|
Country/Territory | India |
City | Kolkata |
Period | 28/03/2022 → 31/03/2022 |
Sponsor | IEEE Engineering in Medicine and Biology Society (EMBS), IEEE Signal Processing Society, Institute of Electrical and Electronic Engineers (IEEE) |
Bibliographical note
Publisher Copyright:© 2022 IEEE.
Keywords
- automated segmentation
- computed tomography
- head and neck radiotherapy planning
- magnetic resonance
- organs-at-risk