Meta multi-task nuclei segmentation with fewer training samples

• Published in: Medical image analysis 2022-08, Vol.80, p.102481-102481, Article 102481
• Main Authors: Han, Chu, Yao, Huasheng, Zhao, Bingchao, Li, Zhenhui, Shi, Zhenwei, Wu, Lei, Chen, Xin, Qu, Jinrong, Zhao, Ke, Lan, Rushi, Liang, Changhong, Pan, Xipeng, Liu, Zaiyi
• Format: Article
• Language: English
• Subjects: Convolutional neural networks; Meta learning; Multi-task learning; Nuclei segmentation
• ISSN: 1361-8415; 1361-8423
• DOI: 10.1016/j.media.2022.102481

•We proposed a meta multi-task learning scheme (Meta-MTL) for nuclei segmentation with better domain adaptation and less data dependency.•In the contour-aware multi-task learning model, we proposed a feature fusion and interaction block (FFIB) for features communication between two parallel tasks.•Our proposed model achieves comparable performances with state-of-the-art models with fewer training samples. And the proposed model performs well on domain adaptation and domain generalization. [Display omitted] 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