Dual-Stream CNN with Graph Neural Network (GNN) Integration for Head and Neck Cancer Recurrence Prediction

DOI

10.2991/978-94-6239-674-6_21How to use a DOI?

Keywords

Head and neck cancer; Dual-stream CNN; Graph neural network; Recurrence prediction; Multimodal fusion; Medical image analysis

Abstract

The recurrence of head and neck cancer is a life-threatening issue in the post-treatment management because of the diverse biology of tumors as well as clinical heterogeneity. In this paper, the authors suggest a Dual-Stream Convolutional Neural Network with a Graph Neural Network as an indicator of recurrence based on CT images and clinical metadata. The model is based on TCIA Head-Neck Radiomics HN1 and HN2 datasets of 495 patients with follow-up outcomes. ResNet-50 and DenseNet-121 branches are applied to primary tumor and lymph- node regions, and clinical variables are presented in the form of a patient similarity graph, which is analyzed with the help of a Graph Attention Network. Cross-attention fusion module is a union of CNN and GNN features that enhances the representation of multimodal features. The experimental findings reveal high levels of performance with an accuracy of 0.912, sensitivity of 0.928 and AUC of 0.937; higher than the single-stream and image-only models. The suggested procedure encourages automatic monitoring and accuracy risk stratification processes that can enhance enhanced decision support in head and neck cancer follow up treatment.

Copyright

© 2026 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Sanjeev Kumar Ojha
AU  - Sujeet Kumar Sahani
AU  - Mohammad Shahrookh Husain
AU  - Dhiresh Kumar Pathak
AU  - Neelam Singh
AU  - Abhishek Varshney
PY  - 2026
DA  - 2026/05/28
TI  - Dual-Stream CNN with Graph Neural Network (GNN) Integration for Head and Neck Cancer Recurrence Prediction
BT  - Proceedings of the International Conference on Sustainable Computing and Artificial Intelligence (ICSCAI 2025)
PB  - Atlantis Press
SP  - 243
EP  - 253
SN  - 2352-5401
UR  - https://doi.org/10.2991/978-94-6239-674-6_21
DO  - 10.2991/978-94-6239-674-6_21
ID  - Ojha2026
ER  -