# Dataset-Aware Cold-Start Active Learning for Annotation-Efficient 3D Medical Image Segmentation

arXiv: https://arxiv.org/abs/2606.20765 · HTML: https://mlsanity.com/paper/2606.20765/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 25 of its 39 references that are in the map.
- First posted on arXiv 18 Jun 2026.

## Builds on (most cited first, 25 of 25)

- [PyTorch: An Imperative Style, High-Performance Deep Learning Library](https://mlsanity.com/paper/1912.01703/) (2019)
- [What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?](https://mlsanity.com/paper/1703.04977/) (2017)
- [Swin UNETR: Swin Transformers for Semantic Segmentation of Brain Tumors in MRI Images](https://mlsanity.com/paper/2201.01266/) (2022)
- [Deep Bayesian Active Learning with Image Data](https://mlsanity.com/paper/1703.02910/) (2017)
- [The Medical Segmentation Decathlon](https://mlsanity.com/paper/2106.05735/) (2021)
- [MONAI: An open-source framework for deep learning in healthcare](https://mlsanity.com/paper/2211.02701/) (2022)
- [Self-Supervised Pre-Training of Swin Transformers for 3D Medical Image Analysis](https://mlsanity.com/paper/2111.14791/) (2021)
- [Variational Adversarial Active Learning](https://mlsanity.com/paper/1904.00370/) (2019)
- [Embracing Imperfect Datasets: A Review of Deep Learning Solutions for Medical Image Segmentation](https://mlsanity.com/paper/1908.10454/) (2019)
- [Metrics reloaded: Recommendations for image analysis validation](https://mlsanity.com/paper/2206.01653/) (2022)
- [Active Learning on a Budget: Opposite Strategies Suit High and Low Budgets](https://mlsanity.com/paper/2202.02794/) (2022)
- [Annotation-efficient deep learning for automatic medical image segmentation](https://mlsanity.com/paper/2012.04885/) (2020)
- [Active Learning Through a Covering Lens](https://mlsanity.com/paper/2205.11320/) (2022)
- [An automatic multi-tissue human fetal brain segmentation benchmark using the Fetal Tissue Annotation Dataset](https://mlsanity.com/paper/2010.15526/) (2020)
- [A comprehensive survey on deep active learning in medical image analysis](https://mlsanity.com/paper/2310.14230/) (2023)
- [Unsupervised Selective Labeling for More Effective Semi-Supervised Learning](https://mlsanity.com/paper/2110.03006/) (2021)
- [Ranger21: a synergistic deep learning optimizer](https://mlsanity.com/paper/2106.13731/) (2021)
- [Warm Start Active Learning with Proxy Labels & Selection via Semi-Supervised Fine-Tuning](https://mlsanity.com/paper/2209.06285/) (2022)
- [On Initial Pools for Deep Active Learning](https://mlsanity.com/paper/2011.14696/) (2020)
- [How to Select Which Active Learning Strategy is Best Suited for Your Specific Problem and Budget](https://mlsanity.com/paper/2306.03543/) (2023)
- [COLosSAL: A Benchmark for Cold-start Active Learning for 3D Medical Image Segmentation](https://mlsanity.com/paper/2307.12004/) (2023)
- [nnActive: A Framework for Evaluation of Active Learning in 3D Biomedical Segmentation](https://mlsanity.com/paper/2511.19183/) (2025)
- [MedCAL-Bench: A Comprehensive Benchmark on Cold-Start Active Learning with Foundation Models for Medical Image Analysis](https://mlsanity.com/paper/2508.03441/) (2025)
- [Finally Outshining the Random Baseline: A Simple and Effective Solution for Active Learning in 3D Biomedical Imaging](https://mlsanity.com/paper/2601.13677/) (2026)
- [From Cold Start to Active Learning: Embedding-Based Scan Selection for Medical Image Segmentation](https://mlsanity.com/paper/2601.18532/) (2026)

## Abstract (arXiv)

Deep learning for 3D medical image segmentation requires extensive manual annotations, a major bottleneck in volumetric medical imaging. Active learning aims to reduce this burden by selecting informative samples for annotation, but most methods assume that an initial labeled set is already available. This leaves the cold-start problem largely unresolved: how to select the first volumes from a fully unlabeled pool before any task-specific model is trained. We propose CSCS, a Curriculum-Stratified Cold-Start framework that adapts initial sample selection to the structure of the unlabeled dataset. CSCS combines two self-supervised, label-free signals: local typicality, measuring representativeness in the embedding space, and reconstruction-based uncertainty, used as a proxy for sample difficulty. These signals are combined through a weighted geometric score, where the weighting is determined by a closed-form pacing rule based on the effective annotation budget and the Difficulty-Coverage Ratio, a pool-level statistic measuring the alignment between difficulty and representativeness. We evaluate CSCS on four 3D medical image segmentation benchmarks: BraTS, FeTA, Spleen, and an in-house fetal MRI dataset. Using nnU-Net as downstream segmentation model, CSCS shows consistently competitive performance across datasets and annotation budgets, with the strongest gains in low-to-mid annotation regimes. These results suggest that dataset-aware cold-start initialization can improve the robustness of active learning for 3D medical image segmentation by adapting sample selection to the geometry of the unlabeled pool.