Comparing the Performance of Foundation Model Derived Embeddings with Traditional Approaches for Distant Metastasis Prediction in Head and Neck Cancer

2026-07-28

AI Summary

This study addresses the challenge of predicting distant metastasis risk in head and neck cancer, a task traditionally reliant on labor-intensive tumor segmentation that is prone to human bias. Leveraging CT imaging, the authors construct three distinct feature sets—radiomic features, deep learning–derived features, and embeddings from a pretrained CT foundation model—and feed each into a multilayer perceptron for prediction. The work systematically evaluates their performance and, for the first time, demonstrates the efficacy of medical foundation model embeddings in this context. This approach eliminates the need for precise tumor delineation, simplifies preprocessing, and achieves an AUC of 0.791, outperforming both radiomics (AUC: 0.772) and deep learning methods (AUC: 0.753), while closely matching the performance of their fused model (AUC: 0.794), thereby offering a clinically efficient and robust new paradigm.

Abstract

Background: Early prediction of distant metastasis (DM) risk in head and neck cancer (HNC) can enable timely interventions that may improve treatment outcomes. Many current machine learning methods rely on prior knowledge of the region of interest such as tumor segmentations, which require expert knowledge, is time-consuming and introduces user-dependent variability. Medical image-based foundation models have recently been developed for specific imaging modalities to streamline down-stream prediction tasks by extracting modality-relevant features. Purpose: In this study, we evaluate the effectiveness of using a foundation model as the feature extractor to predict DM risk in HNC patients and compare its performance with traditional approaches that require prior knowledge on the regions of interest. Methods: Preoperative CT images of 2327 patients from the RADCURE dataset were used. Three features-sets were created including radiomics, deep-learning based features, and CT Foundation derived features. The feature-sets were used individually in a multi-layer perceptron (MLP) to predict DM risk. Results: The model using CT Foundation embeddings outperformed the radiomics and deep learning-based models, achieving a Receiver Operating Characteristic Area Under the Curve (AUC) of 0.791, compared to AUC values of 0.772 and 0.753 for the radiomics and deep learning-based models, respectively. The CT Foundation based model had similar performance to a model that combined the use of radiomics and deep learning-based features that achieved an AUC of 0.794. Conclusions: Features based on foundation models offer a promising alternative to traditional radiomics while reducing the need for domain expertise and extensively annotated datasets. Their minimal preprocessing requirements also make them a more accessible and scalable option.

Problem

• distant metastasis prediction
• head and neck cancer
• foundation model
• radiomics
• medical image analysis

Innovation

• foundation model
• distant metastasis prediction
• head and neck cancer
• radiomics
• medical image embedding