当前位置: X-MOL 学术Methods › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Deep-Hipo: Multi-scale Receptive Field Deep Learning for Histopathological Image Analysis
Methods ( IF 4.2 ) Pub Date : 2020-07-01 , DOI: 10.1016/j.ymeth.2020.05.012
Sai Chandra Kosaraju 1 , Jie Hao 2 , Hyun Min Koh 3 , Mingon Kang 1
Affiliation  

Digitizing whole-slide imaging in digital pathology has led to the advancement of computer-aided tissue examination using machine learning techniques, especially convolutional neural networks. A number of convolutional neural network-based methodologies have been proposed to accurately analyze histopathological images for cancer detection, risk prediction, and cancer subtype classification. Most existing methods have conducted patch-based examinations, due to the extremely large size of histopathological images. However, patches of a small window often do not contain sufficient information or patterns for the tasks of interest. It corresponds that pathologists also examine tissues at various magnification levels, while checking complex morphological patterns in a microscope. We propose a novel multi-task based deep learning model for HIstoPatholOgy (named Deep-Hipo) that takes multi-scale patches simultaneously for accurate histopathological image analysis. Deep-Hipo extracts two patches of the same size in both high and low magnification levels, and captures complex morphological patterns in both large and small receptive fields of a whole-slide image. Deep-Hipo has outperformed the current state-of-the-art deep learning methods. We assessed the proposed method in various types of whole-slide images of the stomach: well-differentiated, moderately-differentiated, and poorly-differentiated adenocarcinoma; poorly cohesive carcinoma, including signet-ring cell features; and normal gastric mucosa. The optimally trained model was also applied to histopathological images of The Cancer Genome Atlas (TCGA), Stomach Adenocarcinoma (TCGA-STAD) and TCGA Colon Adenocarcinoma (TCGA-COAD), which show similar pathological patterns with gastric carcinoma, and the experimental results were clinically verified by a pathologist. The source code of Deep-Hipo is publicly available at http://dataxlab.org/deep-hipo.

中文翻译:

Deep-Hipo:用于组织病理学图像分析的多尺度感受野深度学习

数字病理学中的全载玻片成像数字化促进了使用机器学习技术,尤其是卷积神经网络的计算机辅助组织检查的进步。已经提出了许多基于卷积神经网络的方法来准确分析用于癌症检测、风险预测和癌症亚型分类的组织病理学图像。由于组织病理学图像的尺寸非常大,大多数现有方法都进行了基于补丁的检查。然而,小窗口的补丁通常不包含感兴趣任务的足够信息或模式。相应地,病理学家也在各种放大倍数下检查组织,同时在显微镜下检查复杂的形态学模式。我们为 HIstoPatholOgy(名为 Deep-Hipo)提出了一种新的基于多任务的深度学习模型,该模型同时采用多尺度补丁进行准确的组织病理学图像分析。Deep-Hipo 在高和低放大倍数下提取两个相同大小的块,并在整个幻灯片图像的大小感受野中捕获复杂的形态模式。Deep-Hipo 的性能优于当前最先进的深度学习方法。我们在胃的各种类型的全幻灯片图像中评估了所提出的方法:高分化、中分化和低分化腺癌;粘性差的癌,包括印戒细胞特征;和正常的胃黏膜。优化训练的模型也应用于癌症基因组图谱 (TCGA) 的组织病理学图像,胃腺癌(TCGA-STAD)和TCGA结肠腺癌(TCGA-COAD)的病理模式与胃癌相似,实验结果经病理学家临床验证。Deep-Hipo 的源代码可在 http://dataxlab.org/deep-hipo 上公开获得。
更新日期:2020-07-01
down
wechat
bug