Comparing SAM 2 and SAM 3 for Zero-Shot Segmentation of 3D Medical Data

TL;DR

This study compares SAM 2 and SAM 3 models for zero-shot segmentation of 3D medical data, finding SAM 3 generally outperforms SAM 2 across various modalities and prompting strategies.

Contribution

First controlled comparison of SAM 2 and SAM 3 for 3D medical segmentation, evaluating performance and failure modes across multiple datasets and modalities.

Findings

SAM 3 performs better with click prompts across modalities.

SAM 3 has fewer over-segmentation failures and slower decay of predictions.

Performance differences narrow with bounding-box and mask prompts in some structures.

Abstract

Foundation models, such as the Segment Anything Model (SAM), have heightened interest in promptable zero-shot segmentation. Although these models perform strongly on natural images, their behavior on medical data remains insufficiently characterized. While SAM 2 has been widely adopted for annotation in 3D medical workflows, the recently released SAM 3 introduces a new architecture that may change how visual prompts are interpreted and propagated. Therefore, to assess whether SAM 3 can serve as an out-of-the-box replacement for SAM 2 for zero-shot segmentation of 3D medical data, we present the first controlled comparison of both models by evaluating SAM 3 in its Promptable Visual Segmentation (PVS) mode using a variety of prompting strategies. We benchmark on 16 public datasets (CT, MRI, Ultrasound, endoscopy) covering 54 anatomical structures, pathologies, and surgical instruments. We further quantify three failure modes: prompt-frame over-segmentation, over-propagation after object disappearance, and temporal retention of well-initialized predictions. Our results show that SAM 3 is consistently stronger under click prompting across modalities, with fewer prompt-frame over-segmentation failures and slower prediction retention decay compared to SAM 2. Under bounding-box and mask prompts, performance gaps narrow in few structures of CT/MR and the models trade off termination behavior, while SAM 3 remains stronger on ultrasound and endoscopy sequences. The overall results position SAM 3 as the superior default choice for most medical segmentation tasks, while clarifying when SAM 2 remains a preferable propagator.