Related Experiment Video
Updated: Jan 9, 2026
04:48
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
Published on: November 30, 2022
3.3K
Enhancing Medical Image Segmentation with Adaptive Prompting with Foundation Model in Semi-Supervised Learning
Summary
Adaptive Prompting with Segment Anything Model (ADP-SAM) improves medical image segmentation using unlabeled data. This novel approach enhances accuracy and stability, overcoming limitations of existing semi-supervised learning methods.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Computer Vision
Background:
- Semi-supervised learning is crucial for medical image segmentation due to limited labeled data.
- Conventional methods face instability and inconsistency with unlabeled data, impacting segmentation reliability.
Purpose of the Study:
- To introduce Adaptive Prompting with Segment Anything Model (ADP-SAM) for robust semi-supervised medical image segmentation.
- To enhance segmentation accuracy and stability, especially when labeled data is scarce.
Main Methods:
- Developed ADP-SAM, integrating adaptive prompting and cross-teaching mechanisms.
- Adaptive prompting uses learnable prompts to focus on critical features.
- Cross-teaching employs two models to iteratively refine pseudo-labels for improved consistency.
Main Results:
- ADP-SAM demonstrated superior performance compared to existing semi-supervised techniques on medical image datasets.
- The framework achieved more accurate and stable segmentation outcomes.
- Effectively leveraged unlabeled data for improved segmentation quality.
Conclusions:
- ADP-SAM offers a robust solution for medical image segmentation with limited labeled data.
- The adaptive prompting and cross-teaching mechanisms significantly enhance segmentation performance.
- ADP-SAM shows strong potential for clinical applications requiring precise image segmentation.
