发表机构
XLIM, Université de Poitiers; Université de Limoges, CNRS（普瓦捷大学XLIM实验室; 利摩日大学，法国国家科学研究中心）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
提出Active-DiNTS，将主动学习嵌入双层可微分拓扑搜索，联合优化架构与标注，在单GPU上以少量标注超越现有方法，显著提升3D分割精度与效率。
AI 中文摘要
神经架构搜索（NAS）已被证明是手动网络设计的强有力替代方案，但将其应用于3D医学图像分割受到两个众所周知的成本的限制：大量的标注预算和多GPU集群。因此，本文提出了Active-DiNTS（主动可微分网络拓扑搜索），一种将基于池的主动学习（AL）嵌入双层可微分拓扑搜索的方法，以联合进行架构发现和标签策展。在每个查询轮次中，未标注的MRI体积根据三种不确定性信号（熵、方差或标准差）之一进行排序，只有排名最高的体积被发送给标注者进行标注。新标签反馈到两个相互关联的阶段。网络权重在外循环中更新，而U-Net风格骨干的宏观/微观拓扑在内循环中细化。三种主动学习机制（仅权重、仅拓扑、联合）揭示了速度-精度的权衡。在医学分割十项全能（MSD）Task01脑肿瘤基准上的评估表明，Active-DiNTS在Dice系数上超越了DiNTS、C2FNAS和nnU-Net，在浮肿区域上提升约10个百分点，在非增强核心上提升5个百分点，且仅使用单个GPU和一小部分标注体积。发现的架构比先前基线更密集且FLOP更高，但在可训练参数和峰值内存方面保持竞争力；最快的搜索变体在不到0.25个GPU天内完成，比八GPU的DiNTS搜索快27倍以上。总之，这些结果表明，将可微分NAS与主动数据采集相结合，是在现实约束下实现准确3D分割的实用方案。
英文摘要
Neural Architecture Search (NAS) has proved to be a strong alternative to manual network design, but applying it to 3D medical image segmentation is limited by two well-known costs, large annotation budgets and multi-GPU clusters. Thus, this paper introduces Active-DiNTS (Active Differentiable Network Topology Search), an approach that embeds pool-based Active Learning (AL) into a bi-level differentiable topology search to perform architecture discovery and label curation jointly. At each query round, unlabeled MRI volumes are ranked by one of three uncertainty signals (Entropy, Variance, or Standard Deviation), and only the top-ranked volumes are sent to an oracle for annotation. The new labels feed two interlocked stages. Network weights are updated in an outer loop, while the macro/micro topology of a U-Net-style backbone is refined in an inner loop. Three AL regimes (weights-only, topology-only, joint) expose the speed-accuracy trade-off. Evaluations on the Medical Segmentation Decathlon (MSD) Task01 BrainTumour benchmark showed that Active-DiNTS surpasses DiNTS, C2FNAS, and nnU-Net in Dice, with gains of about 10 percentage points on Edema and 5 points on Non-Enhancing core, on a single GPU and using a fraction of the labeled volumes. The discovered architectures are denser and more FLOP-heavy than prior baselines, but remain competitive in trainable parameters and peak memory; the fastest search variant finishes in under 0.25 GPU-days, over 27x faster than the eight-GPU DiNTS search. Together, these results indicate that pairing differentiable NAS with active data acquisition is a practical recipe for accurate 3D segmentation under realistic constraints.
Comments19 pages, 7 figures, 3 tables. Extends material from the first author's PhD thesis (University of Sao Paulo / La Rochelle Universite, 2024)
