False Negative/Positive Control for SAM on Noisy Medical Images
This work addresses the problem of adapting foundation models like SAM to noisy medical images for practitioners, but it is incremental as it refines existing techniques without major methodological breakthroughs.
The paper tackled SAM's poor segmentation performance on noisy, low-contrast medical images like ultrasound by proposing a test-phase prompt augmentation and false-negative/positive correction method, which improved SAM's performance and robustness without retraining, as shown on two ultrasound datasets.
The Segment Anything Model (SAM) is a recently developed all-range foundation model for image segmentation. It can use sparse manual prompts such as bounding boxes to generate pixel-level segmentation in natural images but struggles in medical images such as low-contrast, noisy ultrasound images. We propose a refined test-phase prompt augmentation technique designed to improve SAM's performance in medical image segmentation. The method couples multi-box prompt augmentation and an aleatoric uncertainty-based false-negative (FN) and false-positive (FP) correction (FNPC) strategy. We evaluate the method on two ultrasound datasets and show improvement in SAM's performance and robustness to inaccurate prompts, without the necessity for further training or tuning. Moreover, we present the Single-Slice-to-Volume (SS2V) method, enabling 3D pixel-level segmentation using only the bounding box annotation from a single 2D slice. Our results allow efficient use of SAM in even noisy, low-contrast medical images. The source code will be released soon.