Meta multi-task nuclei segmentation with fewer training samples
Meta multi-task nuclei segmentation with fewer training samples
复制标题
DOI：
10.1016/j.media.2022.102481
复制
发表时间：
2022-05-30
影响因子：
10.9
通讯作者：
Liu, Zaiyi
中科院分区：
文献类型：
--
作者：
Han, Chu;Yao, Huasheng;Liu, Zaiyi
细胞/核传递着大量的微环境信息。一种自动的细胞核分割方法可以减少病理学家的工作量，并为生物学和临床研究提供精确的微环境。现有的深度学习模型在大量标注数据的监督下取得了优异的性能。然而，当来自看不见的领域的数据到来时，我们仍然需要准备一定程度的手动注释，以便为每个领域进行训练。不幸的是，获取组织病理学注释极其困难。它高度依赖专业知识，而且非常耗时。在本文中，我们试图建立一个数据依赖性更小、更具泛化能力的广义核分割模型。为此，我们提出了一种需要较少训练样本的核分割元多任务学习(Meta-MTL)模型。模型不可知元学习被用作分割模型的外部优化算法。我们引入了一个轮廓感知的多任务学习模型作为内部模型。提出了一种特征融合和交互模块(FFIB)，以允许跨两个任务的特征通信。大量实验证明，我们提出的Meta-MTL模型可以提高模型的泛化能力，并在训练样本较少的情况下获得与现有模型相当的性能。我们的模型也可以在看不见的领域上进行快速的自适应，只需少量的人工注释。代码可在https://github.com/ChuHan89/Meta-MTL4NucleiSegmentation(C)2022 Elsevier B.V.获得。保留所有权利。
Cells/nuclei deliver massive information of microenvironment. An automatic nuclei segmentation approach can reduce pathologists' workload and allow precise of the microenvironment for biological and clinical researches. Existing deep learning models have achieved outstanding performance under the supervision of a large amount of labeled data. However, when data from the unseen domain comes, we still have to prepare a certain degree of manual annotations for training for each domain. Unfortunately, obtaining histopathological annotations is extremely difficult. It is high expertise-dependent and time-consuming. In this paper, we attempt to build a generalized nuclei segmentation model with less data dependency and more generalizability. To this end, we propose a meta multi-task learning (Meta-MTL) model for nuclei segmentation which requires fewer training samples. A model-agnostic meta-learning is applied as the outer optimization algorithm for the segmentation model. We introduce a contour-aware multi-task learning model as the inner model. A feature fusion and interaction block (FFIB) is proposed to allow feature communication across both tasks. Extensive experiments prove that our proposed Meta-MTL model can improve the model generalization and obtain a comparable performance with state-of-the-art models with fewer training samples. Our model can also perform fast adaptation on the unseen domain with only a few manual annotations. Code is available at https://github.com/ChuHan89/Meta-MTL4NucleiSegmentation (c) 2022 Elsevier B.V. All rights reserved.