Abstract

Radiotherapy is a critical cancer treatment that uses external beam radiation to kill cancer cells. Effective treatment planning, which determines radiation dose distribution for tumor volumes, is essential to maximize the likelihood of a cure while minimizing toxicity. Gross Tumor Volume (GTV) and Clinical Target Volume (CTV) are two key targets in radiotherapy planning. GTV is the visible tumor extent on Computed Tomography (CT) scans, while CTV encompasses GTV and describes the extent of microscopic, unimageable tumor spread. Accurate delineation of GTV and CTV is essential in radiotherapy planning, however, manual slice-by-slice annotation on CT scans is time-consuming and labor-intensive for radiation oncologists. Automating this process can significantly reduce planning time and enhance radiotherapy efficiency.
Based on the success of SegRap2023, SegRap2025 aims to address ongoing challenges in GTV segmentation and firstly focus on the Lymph Node (LN) CTV segmentation. It provides two datasets collected from diverse cohorts: one with data from 260 NPC patients annotated for GTV, and another with 402 patients annotated for LN CTV. It also provides an unlabeled dataset of CT scans from 500 patients to support model training. Based on the extensive and comprehensive datasets, two sub-tasks will be held in SegRap2025:
• Task01: GTV Segmentation
• Task02: LN CTV Segmentation
We believe that these algorithms will have the potential to support clinical manual delineation, enhance research on radiation dose calculation, and improve efficiency of radiotherapy treatment planning.

Important Dates

• Registration opens: May 1st (12:00 AM GMT), 2025
• Release of training data: May 10th (12:00 AM GMT), 2025
• Release of validation data: June 30th (12:00 AM GMT), 2025
• Docker and short paper submission opens: Aug. 10th (12:00 AM GMT), 2025
• Submission deadline: Aug. 31st (12:00 AM GMT), 2025
• Announcement of final results: Sep. 23th and 27th, 2025

References

You can find our previous challenges here: SegRap2023, associated paper with summary of the results can be downloaded here, and algorithms from top teams can be found at: Google-Drive.

Please cite the following when using the SegRap dataset in your research:
• X. Luo et al. Segrap2023: A benchmark of organs-at-risk and gross tumor volume segmentation for radiotherapy planning of nasopharyngeal carcinoma. Medical image analysis 2025, 101: 103447.
• X. Luo et al. A multicenter dataset for lymph node clinical target volume delineation of nasopharyngeal carcinoma. Scientific Data 11, 1085 (2024).