用MRI标注训练模型，实现术中超声脑肿瘤自动分割。
Automatic brain tumor segmentation in 2D intra-operative ultrasound images using magnetic resonance imaging tumor annotations
- 用MRI标注迁移至超声图像，替代稀缺的超声标注训练模型。
- 模型平均Dice达0.62，与神经外科医生水平相当（0.67）。
- 剔除小肿瘤可提升性能，适合医疗影像自动化研究者。
术中超声（iUS）图像中脑肿瘤的自动分割有助于手术切除时定位肿瘤组织。当前模型性能受限于缺乏大规模标注数据集。本文研究利用更易获取的磁共振成像（MRI）肿瘤标注，训练深度学习模型以实现iUS脑肿瘤分割。采用180例带标注的MRI扫描及对应未标注的iUS图像，以及29例带标注的iUS图像。通过图像配准将MRI标注转移到相应iUS图像后，使用nnU-Net模型在不同数据与标签来源配置下进行训练。结果表明，仅用MRI标注训练的模型与仅用iUS标注或两者结合训练的模型，以及专家标注相比，其Dice分数无显著差异，说明MRI标注可作为iUS标注的有效替代。最佳模型平均Dice为0.62±0.31，优于专家神经外科医生的0.67±0.25；对大肿瘤表现相近，但对小肿瘤性能较低。此外，移除训练集中小肿瘤可提升效果。主要模型代码已公开：https://github.com/mathildefaanes/us_brain_tumor_segmentation/tree/main。
原文摘要 · Abstract (English)
Automatic segmentation of brain tumors in intra-operative ultrasound (iUS) images could facilitate localization of tumor tissue during resection surgery. The lack of large annotated datasets limits the current models performances. In this paper, we investigated the use of tumor annotations in magnetic resonance imaging (MRI) scans, which are more accessible than annotations in iUS images, for training of deep learning models for iUS brain tumor segmentation. We used 180 annotated MRI scans with corresponding unannotated iUS images, and 29 annotated iUS images. Image registration was performed to transfer the MRI annotations to the corresponding iUS images before training the nnU-Net model with different configurations of the data and label origins. The results showed no significant difference in Dice score for a model trained with only MRI annotated tumors compared to models trained with only iUS annotations and both, and to expert annotations, indicating that MRI tumor annotations can be used as a substitute for iUS tumor annotations to train a deep learning model for automatic brain tumor segmentation in iUS images. The best model obtained an average Dice score of $0.62\pm0.31$, compared to $0.67\pm0.25$ for an expert neurosurgeon, where the performance on larger tumors were similar, but lower for the models on smaller tumors. In addition, the results showed that removing smaller tumors from the training sets improved the results. The main models are available here: https://github.com/mathildefaanes/us_brain_tumor_segmentation/tree/main
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。