Multi-Task Deep Learning for Head and Neck Cancer: Segmentation, Survival Prediction, and HPV Classification in the HECKTOR 2025 Challenge

Jakob Dexl, Michael Ingrisch

Keywords: Head and neck cancer, PET/CT, Segmentation, FDG

TL;DR: This report presents the submissions of Team CDS to the HECKTOR 2025 challenge.

Abstract:

This paper describes our submissions (team CDS) to the
HECKTOR 2025 challenge, which addresses three tasks: (1) tumor and
lymph node segmentation, (2) recurrence-free survival prediction, and
(3) HPV status classification. For Task 1, we trained a baseline UNet
and refined the final model using stochastic weight averaging and small
lesion removal. For Task 2, we employed a lightweight 3D ResNet18
that combines PET, CT, segmentation masks, and clinical metadata,
optimized with a Cox loss. For Task 3, we extended the segmenta-
tion model with a classification head and metadata integration. Cross-
validation results were promising, performance on the preliminary vali-
dation set was however lower, underlining the challenges of generaliza-
tion in multi-center cohorts. Code and trained models are available at
github.com/JakobDexl/HECKTOR25

Submission Number: 12