Can Foundation Models Really Segment Tumors? A Benchmarking Odyssey in Lung CT Imaging

TL;DR

This paper benchmarks various deep learning models, including foundation models, for lung tumor segmentation in CT images, demonstrating that foundation models like MedSAM~2 outperform traditional methods in accuracy and efficiency.

Contribution

It provides a comprehensive comparison of traditional, self-configuring, and foundation models for lung tumor segmentation, highlighting the superior performance of foundation models.

Findings

Foundation models outperform traditional models in accuracy.

MedSAM~2 achieves the best segmentation performance.

Foundation models are more computationally efficient.

Abstract

Accurate lung tumor segmentation is crucial for improving diagnosis, treatment planning, and patient outcomes in oncology. However, the complexity of tumor morphology, size, and location poses significant challenges for automated segmentation. This study presents a comprehensive benchmarking analysis of deep learning-based segmentation models, comparing traditional architectures such as U-Net and DeepLabV3, self-configuring models like nnUNet, and foundation models like MedSAM, and MedSAM~2. Evaluating performance across two lung tumor segmentation datasets, we assess segmentation accuracy and computational efficiency under various learning paradigms, including few-shot learning and fine-tuning. The results reveal that while traditional models struggle with tumor delineation, foundation models, particularly MedSAM~2, outperform them in both accuracy and computational efficiency. These findings underscore the potential of foundation models for lung tumor segmentation, highlighting their applicability in improving clinical workflows and patient outcomes.