The construction of 3D medical image datasets presents several issues, including requiring significant financial costs in data collection and specialized expertise for annotation, as well as strict privacy concerns for patient confidentiality compared to natural image datasets. Therefore, it has become a pressing issue in 3D medical image segmentation to enable data-efficient learning with limited 3D medical data and supervision. A promising approach is pre-training, but improving its performance in 3D medical image segmentation is difficult due to the small size of existing 3D medical image datasets. We thus present the Primitive Geometry Segment Pre-training (PrimGeoSeg) method to enable the learning of 3D semantic features by pre-training segmentation tasks using only primitive geometric objects for 3D medical image segmentation. PrimGeoSeg performs more accurate and efficient 3D medical image segmentation without manual data collection and annotation. Further, experimental results show that PrimGeoSeg on SwinUNETR improves performance over learning from scratch on BTCV, MSD (Task06), and BraTS datasets by 3.7%, 4.4%, and 0.3%, respectively. Remarkably, the performance was equal to or better than state-of-the-art self-supervised learning despite the equal number of pre-training data. From experimental results, we conclude that effective pre-training can be achieved by looking at primitive geometric objects only. Code and dataset are available at https://github.com/SUPER-TADORY/PrimGeoSeg.

翻译：3D医学图像数据集构建面临多重挑战：与传统自然图像数据集相比，其数据收集需要高昂财务成本，标注需要专业知识，且患者隐私保护面临严格限制。因此，如何在有限3D医学数据和监督条件下实现数据高效学习，已成为3D医学图像分割领域亟待解决的问题。预训练方法虽具潜力，但由于现有3D医学图像数据集规模较小，其在提升分割性能方面仍存在困难。为此，我们提出原始几何分割预训练方法（PrimGeoSeg），通过仅使用原始几何对象进行分割任务预训练，实现3D语义特征学习，从而完成3D医学图像分割。该方法无需人工数据收集与标注，即可实现更精准高效的3D医学图像分割。实验结果表明，在SwinUNETR架构上应用PrimGeoSeg预训练，相较于从头训练，在BTCV、MSD（Task06）和BraTS数据集上性能分别提升3.7%、4.4%和0.3%。值得注意的是，在预训练数据量相等的情况下，该方法性能达到甚至超越当前最优自监督学习方法。基于实验结果，我们得出结论：仅通过观察原始几何对象即可实现有效的预训练。代码与数据集已开源至https://github.com/SUPER-TADORY/PrimGeoSeg。