首次系统研究3D医学目标检测预训练，发现重建自监督效果最佳。
The Missing Piece: A Case for Pre-Training in 3D Medical Object Detection
- 对比多种预训练方法在3D检测架构中的表现
- 重建自监督预训练显著提升检测性能，优于有监督方法
- 适合关注医学影像检测的算法研究者和临床辅助诊断开发者
大规模预训练有望推动3D医学目标检测的发展，这是精准计算机辅助诊断的关键环节。然而，与已证明预训练显著有益的分割任务相比，该领域仍研究不足。现有3D目标检测预训练方法多依赖2D医学数据或自然图像预训练，未能充分利用3D体数据信息。本文首次系统研究将现有预训练方法整合到先进检测架构（包括CNN与Transformer）中的效果。结果表明，预训练在多个任务与数据集上均一致提升检测性能。值得注意的是，基于重建的自监督预训练优于有监督预训练，而对比学习预训练对3D医学目标检测无明显增益。代码已公开于：https://github.com/MIC-DKFZ/nnDetection-finetuning。
原文摘要 · Abstract (English)
Large-scale pre-training holds the promise to advance 3D medical object detection, a crucial component of accurate computer-aided diagnosis. Yet, it remains underexplored compared to segmentation, where pre-training has already demonstrated significant benefits. Existing pre-training approaches for 3D object detection rely on 2D medical data or natural image pre-training, failing to fully leverage 3D volumetric information. In this work, we present the first systematic study of how existing pre-training methods can be integrated into state-of-the-art detection architectures, covering both CNNs and Transformers. Our results show that pre-training consistently improves detection performance across various tasks and datasets. Notably, reconstruction-based self-supervised pre-training outperforms supervised pre-training, while contrastive pre-training provides no clear benefit for 3D medical object detection. Our code is publicly available at: https://github.com/MIC-DKFZ/nnDetection-finetuning.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。