Breaking the Barrier: Selective Uncertainty-based Active Learning for Medical Image Segmentation
This work addresses the challenge of reducing annotation workload and enhancing performance in medical image segmentation, particularly for imbalanced settings like lesion detection, though it is incremental as it builds on existing uncertainty-based methods.
The paper tackled the problem of conventional uncertainty-based active learning methods neglecting target regions and introducing redundancy in medical image segmentation, resulting in a novel approach that prioritizes pixels in target areas and near decision boundaries, achieving substantial improvements across five methods and two datasets with fewer labeled data.
Active learning (AL) has found wide applications in medical image segmentation, aiming to alleviate the annotation workload and enhance performance. Conventional uncertainty-based AL methods, such as entropy and Bayesian, often rely on an aggregate of all pixel-level metrics. However, in imbalanced settings, these methods tend to neglect the significance of target regions, eg., lesions, and tumors. Moreover, uncertainty-based selection introduces redundancy. These factors lead to unsatisfactory performance, and in many cases, even underperform random sampling. To solve this problem, we introduce a novel approach called the Selective Uncertainty-based AL, avoiding the conventional practice of summing up the metrics of all pixels. Through a filtering process, our strategy prioritizes pixels within target areas and those near decision boundaries. This resolves the aforementioned disregard for target areas and redundancy. Our method showed substantial improvements across five different uncertainty-based methods and two distinct datasets, utilizing fewer labeled data to reach the supervised baseline and consistently achieving the highest overall performance. Our code is available at https://github.com/HelenMa9998/Selective\_Uncertainty\_AL.