Annotation-efficient training of medical image segmentation network based on scribble guidance in difficult areas
Annotation-efficient training of medical image segmentation network based on scribble guidance in difficult areas
复制标题
DOI：
10.1007/s11548-023-02931-0
复制
发表时间：
2023-05
影响因子：
3
通讯作者：
Mingrui Zhuang;Zhonghua Chen;Yuxin Yang;Lauri Kettunen;Hongkai Wang
中科院分区：
文献类型：
--
作者：
Mingrui Zhuang;Zhonghua Chen;Yuxin Yang;Lauri Kettunen;Hongkai Wang
目的深度医学图像分割网络的训练通常需要大量的人工标注数据。为了减轻人力劳动的负担，人们开发了许多半监督或无监督的方法。然而，由于临床场景的复杂性，训练标签的不足仍然会导致在异质肿瘤、边界模糊等局部困难区域分割不准确。方法提出了一种标注高效的训练方法，只需要在困难的地方进行潦草的指导。分割网络首先用少量的完全标注的数据进行训练，然后用更多的训练数据生成伪标签。人类监督者在不正确的伪标签区域（即困难区域）上涂鸦，并且使用概率调制的测地线变换将涂鸦转换为伪标签地图。为了减少伪标签中潜在误差的影响，通过综合考虑像素到涂鸦的测地线距离和网络输出概率，生成伪标签的置信度图。随着网络的更新，对伪标签和置信图进行迭代优化，伪标签和置信图依次促进网络的训练。结果基于两个数据集（脑肿瘤MRI和肝脏肿瘤CT）的交叉验证表明，我们的方法在保持困难区域（如肿瘤）分割精度的同时显著缩短了标注时间。使用90张潦草标注的训练图像（标注时间：~ 9小时），我们的方法获得了与使用45张完全标注的图像（标注时间：100小时）相同的性能，但所需的标注时间要短得多。结论与传统的全标注方法相比，该方法将人工监督集中在最困难的区域，显著节省了标注工作量。它为复杂临床场景下医学图像分割网络的训练提供了一种高效的标注方法。
PurposeThe training of deep medical image segmentation networks usually requires a large amount of human-annotated data. To alleviate the burden of human labor, many semi- or non-supervised methods have been developed. However, due to the complexity of clinical scenario, insufficient training labels still causes inaccurate segmentation in some difficult local areas such as heterogeneous tumors and fuzzy boundaries.MethodsWe propose an annotation-efficient training approach, which only requires scribble guidance in the difficult areas. A segmentation network is initially trained with a small amount of fully annotated data and then used to produce pseudo labels for more training data. Human supervisors draw scribbles in the areas of incorrect pseudo labels (i.e., difficult areas), and the scribbles are converted into pseudo label maps using a probability-modulated geodesic transform. To reduce the influence of the potential errors in the pseudo labels, a confidence map of the pseudo labels is generated by jointly considering the pixel-to-scribble geodesic distance and the network output probability. The pseudo labels and confidence maps are iteratively optimized with the update of the network, and the network training is promoted by the pseudo labels and the confidence maps in turn.ResultsCross-validation based on two data sets (brain tumor MRI and liver tumor CT) showed that our method significantly reduces the annotation time while maintains the segmentation accuracy of difficult areas (e.g., tumors). Using 90 scribble-annotated training images (annotated time: ~ 9 h), our method achieved the same performance as using 45 fully annotated images (annotation time: > 100 h) but required much shorter annotation time.ConclusionCompared to the conventional full annotation approaches, the proposed method significantly saves the annotation efforts by focusing the human supervisions on the most difficult regions. It provides an annotation-efficient way for training medical image segmentation networks in complex clinical scenario.