Robust Multicentre Detection and Classification of Colorectal Liver Metastases on CT: Application of Foundation Models

2026-01-12

AI Summary

This study addresses the challenge of detecting and classifying colorectal cancer liver metastases (CRLM) in multicenter CT imaging, where data heterogeneity poses significant obstacles. The work proposes a unified framework that, for the first time, integrates the medical foundation model UMedPT with a multi-layer perceptron (MLP) classification head and an FCOS detection head. The framework further incorporates uncertainty quantification and Grad-CAM-based interpretability mechanisms. Evaluated on multicenter datasets from EuCanImage and TCIA, the method achieves a patient-level classification AUC of 0.90 (sensitivity: 0.82), which improves to 0.91 after excluding the 20% most uncertain samples. At the lesion level, it attains an overall detection rate of 69.1%, with detection performance for small-to-large lesions markedly increasing from 30% to 98%, demonstrating robustness and strong clinical applicability.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Computer Vision: Multi-modal Vision
• Reasoning under Uncertainty: Uncertainty Representations

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Colorectal liver metastases (CRLM) are a major cause of cancer-related mortality, and reliable detection on CT remains challenging in multi-centre settings. We developed a foundation model-based AI pipeline for patient-level classification and lesion-level detection of CRLM on contrast-enhanced CT, integrating uncertainty quantification and explainability. CT data from the EuCanImage consortium (n=2437) and an external TCIA cohort (n=197) were used. Among several pretrained models, UMedPT achieved the best performance and was fine-tuned with an MLP head for classification and an FCOS-based head for lesion detection. The classification model achieved an AUC of 0.90 and a sensitivity of 0.82 on the combined test set, with a sensitivity of 0.85 on the external cohort. Excluding the most uncertain 20 percent of cases improved AUC to 0.91 and balanced accuracy to 0.86. Decision curve analysis showed clinical benefit for threshold probabilities between 0.30 and 0.40. The detection model identified 69.1 percent of lesions overall, increasing from 30 percent to 98 percent across lesion size quartiles. Grad-CAM highlighted lesion-corresponding regions in high-confidence cases. These results demonstrate that foundation model-based pipelines can support robust and interpretable CRLM detection and classification across heterogeneous CT data.

Problem

• colorectal liver metastases
• CT imaging
• multicentre detection
• lesion classification
• medical image analysis

Innovation

• foundation model
• uncertainty quantification
• explainable AI
• multicentre generalization
• lesion detection