发表机构
University of Electronic Science and Technology of China; Glassgow College, University of Electronic Science and Technology of China; Shanghai Artificial Intelligence Laboratory; West China Hospital, Sichuan University（电子科技大学; 电子科技大学格拉斯哥学院; 上海人工智能实验室; 四川大学华西医院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
针对医学图像分割中SAM冷启动主动适应的样本选择问题，提出SUGFW+框架，集成SAM的PFUC、PGDR模块与GSCU策略及UPFT过程，在四个公共数据集上取得CSAL任务的最优性能。
AI 中文摘要
冷启动主动学习（CSAL）在低标注预算下，通过从未标注训练集中查询小子集进行标注来提升医学图像分割模型的性能，具有重要意义。现有CSAL方法通常依赖低效的、特定于数据集的自监督学习（SSL），将未标注图像映射到特征空间以进行样本选择。最近，Segment Anything Model（SAM）等基础模型的出现提供了一种有前景的替代方案，因为预训练模型可提供强泛化性特征嵌入，微调（适应）后在下游任务中表现出色。然而，如何在低标注预算的适应过程中系统利用SAM的固有嵌入进行冷启动样本选择，仍未得到充分探索。为解决该问题，我们提出一种扩展的基于SAM的不确定性引导特征加权（SUGFW+）框架，用于SAM的CSAL和适应。具体而言，该框架利用SAM进行补丁级特征与不确定性计算（PFUC），并引入补丁级全局区分表示（PGDR）模块，将补丁级嵌入聚合为高区分性、感知不确定性的图像级特征。随后，这些特征被用于基于聚类与不确定性的贪心选择（GSCU）策略，在样本选择中结合多样性与不确定性。与先前将样本选择与模型训练解耦的CSAL方法不同，SUGFW+通过SAM的不确定性提示微调（UPFT）过程，将这两个阶段紧密集成。在四个公共数据集上的大量实验表明，SUGFW+相较于现有CSAL方法达到了最先进的性能。代码可在此https URL获取。
英文摘要
Cold Start Active Learning (CSAL) is important in improving the performance of a medical image segmentation model with low annotation budget by querying a small subset for annotation from an unlabeled training set. Existing CSAL methods typically rely on inefficient dataset-specific Self-Supervised Learning (SSL) to map the unlabeled images into a feature space for sample selection. Recently, the advent of foundation models such as the Segment Anything Model (SAM) offer a promising alternative as the pre-trained model can provide strong generalizable feature embeddings, and allow high performance in downstream tasks after fine-tuning (adaptation). However, how to systematically exploit SAM's inherent embeddings for cold-start sample selection during adaptation with low annotation budget remains underexplored. To address this, we propose an extended SAM-based Uncertainty-guided Feature Weighting (SUGFW+) framework for CSAL and adaptation of SAM. Specifically, it leverages the SAM for Patch-level Feature and Uncertainty Calculation (PFUC), and introduces a Patch-based Global Distinct Representation (PGDR) module that aggregates patch-level embeddings into highly discriminative, uncertainty-aware image-level features. These features are then utilized by a Greedy Selection with Cluster and Uncertainty (GSCU) strategy to combine diversity and uncertainty during sample selection. Unlike prior CSAL methods that decouple sample selection from model training, SUGFW+ tightly integrates these two stages via an Uncertainty-Prompted Fine-Tuning (UPFT) process of SAM in model training. Extensive experiments on four public datasets demonstrate that SUGFW+ achieves state-of-the-art performance against existing CSAL methods. Code is available at https://github.com/HiLab-git/SUGFW-plus.
