Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Point-supervised Brain Tumor Segmentation with Box-prompted MedSAM

Xiaofeng Liu, Jonghye Woo, Chao Ma, Jinsong Ouyang, Georges El Fakhri

TL;DR

This work addresses the challenge of point-supervised brain tumor segmentation by introducing an iterative framework that leverages a box-prompt MedSAM. It combines a semantic box-prompt generator (SBPG) with a prompt-guided spatial refinement (PGSR), and trains a lightweight prompt refiner to imbue semantic awareness without fine-tuning the foundation model. The method forms a class prototype from memory to score candidate box proposals and refines the segmentation across multiple rounds, achieving substantial gains over point-based baselines and approaching box-supervised MedSAM performance on BraTS2018 whole-tumor segmentation. The results demonstrate the practical potential of using modular vision foundation models to improve weakly supervised medical image segmentation for image-guided interventions.

Abstract

Delineating lesions and anatomical structure is important for image-guided interventions. Point-supervised medical image segmentation (PSS) has great potential to alleviate costly expert delineation labeling. However, due to the lack of precise size and boundary guidance, the effectiveness of PSS often falls short of expectations. Although recent vision foundational models, such as the medical segment anything model (MedSAM), have made significant advancements in bounding-box-prompted segmentation, it is not straightforward to utilize point annotation, and is prone to semantic ambiguity. In this preliminary study, we introduce an iterative framework to facilitate semantic-aware point-supervised MedSAM. Specifically, the semantic box-prompt generator (SBPG) module has the capacity to convert the point input into potential pseudo bounding box suggestions, which are explicitly refined by the prototype-based semantic similarity. This is then succeeded by a prompt-guided spatial refinement (PGSR) module that harnesses the exceptional generalizability of MedSAM to infer the segmentation mask, which also updates the box proposal seed in SBPG. Performance can be progressively improved with adequate iterations. We conducted an evaluation on BraTS2018 for the segmentation of whole brain tumors and demonstrated its superior performance compared to traditional PSS methods and on par with box-supervised methods.

Point-supervised Brain Tumor Segmentation with Box-prompted MedSAM

TL;DR

This work addresses the challenge of point-supervised brain tumor segmentation by introducing an iterative framework that leverages a box-prompt MedSAM. It combines a semantic box-prompt generator (SBPG) with a prompt-guided spatial refinement (PGSR), and trains a lightweight prompt refiner to imbue semantic awareness without fine-tuning the foundation model. The method forms a class prototype from memory to score candidate box proposals and refines the segmentation across multiple rounds, achieving substantial gains over point-based baselines and approaching box-supervised MedSAM performance on BraTS2018 whole-tumor segmentation. The results demonstrate the practical potential of using modular vision foundation models to improve weakly supervised medical image segmentation for image-guided interventions.

Abstract

Delineating lesions and anatomical structure is important for image-guided interventions. Point-supervised medical image segmentation (PSS) has great potential to alleviate costly expert delineation labeling. However, due to the lack of precise size and boundary guidance, the effectiveness of PSS often falls short of expectations. Although recent vision foundational models, such as the medical segment anything model (MedSAM), have made significant advancements in bounding-box-prompted segmentation, it is not straightforward to utilize point annotation, and is prone to semantic ambiguity. In this preliminary study, we introduce an iterative framework to facilitate semantic-aware point-supervised MedSAM. Specifically, the semantic box-prompt generator (SBPG) module has the capacity to convert the point input into potential pseudo bounding box suggestions, which are explicitly refined by the prototype-based semantic similarity. This is then succeeded by a prompt-guided spatial refinement (PGSR) module that harnesses the exceptional generalizability of MedSAM to infer the segmentation mask, which also updates the box proposal seed in SBPG. Performance can be progressively improved with adequate iterations. We conducted an evaluation on BraTS2018 for the segmentation of whole brain tumors and demonstrated its superior performance compared to traditional PSS methods and on par with box-supervised methods.
Paper Structure (4 sections, 2 equations, 1 figure, 1 table)

This paper contains 4 sections, 2 equations, 1 figure, 1 table.

Table of Contents

  1. Introduction
  2. Methodology
  3. Experiments and Results
  4. Conclusion

Figures (1)

  • Figure 1: Proposed iterative refinement framework with SBPG and PGSR modules for semantic-aware PSS utilizing the off-the-shelf box-prompt MedSAM. Only prompt refiner $\theta$ is to be trained with point-supervision.