ESTRO 2024 - Abstract Book

S3120

Physics - Autosegmentation

ESTRO 2024

Conclusion:

In accordance with clinical experience, we designed a network architecture that can integrate routinely available CECT information for high performance CTV delineation for prostate cancer patients. Quantitative evaluations demonstrates that superior CTV delineation is achievable by seamlessly amalgamating CECT data via the Onet. The proposed method significantly boosts the accuracy of CTV delineation in various dimensions, and it allows us to fasten the challenging and labor-intensive process with consistent delineation outcome.

Keywords: contrast-enhanced CT,automatic delineation

2643

Mini-Oral

Brain metastases segmentation using deep learning with spatial information of brain parcellation