SGP-SAM: Self-Gated Prompting for Transferring 3D Segment Anything Models to Lesion Segmentation
This work improves lesion segmentation in 3D medical images for clinicians, but the gains are incremental as it builds on existing SAM-Med3D.
SGP-SAM introduces a self-gated prompting framework to transfer 3D SAM to lesion segmentation, addressing weak spatial representation and class imbalance. It achieves a 7.3% mDice improvement on MSD Liver Tumor over fine-tuning baselines.
Large segmentation foundation models such as the Segment Anything Model (SAM) have reshaped promptable segmentation in natural images, and recent efforts have extended these models to medical images and volumetric settings. However, directly transferring a 3D SAM-style model to lesion segmentation remains challenging due to (i) weak spatial representational capacity for small, irregular targets in intermediate features, and (ii) extreme foreground-background imbalance in 3D volumes.We propose SGP-SAM, a self-gated prompting framework for efficient and effective transfer to 3D lesion segmentation. Our key component, the Self-Gated Prompting Module (SGPM), performs conditional multi-scale spatial enhancement: a lightweight multi-channel gating unit predicts whether the current features require additional multi-scale fusion, and only then activates a Multi-Scale Feature Fusion Block to enrich spatial context. To further address small-lesion learning, we design a Zoom Loss that up-weights lesion-focused supervision by combining Dice and a voxel-balanced focal term.Experiments on MSD Liver Tumor and MSD Brain Tumor (enhancing tumor) show consistent gains over strong transfer baselines based on SAM-Med3D. On MSD Liver Tumor, SGP-SAM improves mDice by 7.3% over fine-tuning.