Digital synthesis of histological stains using micro-structured and multiplexed virtual staining of label-free tissue
Digital synthesis of histological stains using micro-structured and multiplexed virtual staining of label-free tissue
复制标题
DOI：
10.1038/s41377-020-0315-y
复制
发表时间：
2020-05-06
影响因子：
19.4
通讯作者：
Ozcan, Aydogan
中科院分区：
文献类型：
--
作者：
Zhang, Yijie;de Haan, Kevin;Ozcan, Aydogan
组织学染色是诊断各种疾病的重要步骤，并且已经使用了超过世纪来提供组织切片中的对比度，使得组织成分可见以供医学专家进行显微镜分析。然而，这个过程是耗时的，劳动密集型的，昂贵的和破坏性的标本。最近，已经使用组织染色特异性深度神经网络证明了对未标记的组织切片进行虚拟染色的能力，完全避免了组织化学染色步骤。在这里，我们提出了一个新的基于深度学习的框架，该框架使用无标签组织图像生成虚拟染色图像，其中不同的染色剂根据用户定义的微结构图进行合并。这种方法使用单个深度神经网络，该网络接收两个不同的信息源作为其输入：（1）无标记组织样本的自发荧光图像和（2）“数字染色矩阵”，其表示要在同一组织切片中虚拟生成的不同染色剂的所需显微镜图。该数字染色矩阵还可用于虚拟混合现有染色剂，数字合成新的组织学染色剂。我们使用未标记的肾组织切片训练并盲测该虚拟染色网络，以生成苏木精和伊红（H&E）、琼斯银染色和马森三色染色的微结构组合。该方法使用单个网络，将无标记组织图像的虚拟染色与多种类型的染色进行复用，并为合成可在相同组织横截面中创建的新数字组织学染色铺平了道路，这在目前标准组织化学染色方法中是不可行的。神经网络：一种机器学习方法使用多种数字染色剂的组合来突出组织样本中的微观元素，避免了延迟，不一致，有时需要多次活检，所有这些都是传统手工组织染色技术的特征。基于深度学习的组织染色框架由洛杉矶加州大学的Aydogan Ozcan，Yair Rivenson及其同事开发。他们使用自己的方法来训练神经网络，用三种不同类型的染料之一或其组合对肾脏样本进行虚拟染色。与手动染色的组织样品的比较表明，虚拟染色是高度准确的。使用单个神经网络在同一组织样本上使用多个数字染色和虚拟染色混合的能力可以使病理学家从组织中获得更多相关信息，从而改善诊断。
Histological staining is a vital step in diagnosing various diseases and has been used for more than a century to provide contrast in tissue sections, rendering the tissue constituents visible for microscopic analysis by medical experts. However, this process is time consuming, labour intensive, expensive and destructive to the specimen. Recently, the ability to virtually stain unlabelled tissue sections, entirely avoiding the histochemical staining step, has been demonstrated using tissue-stain-specific deep neural networks. Here, we present a new deep-learning-based framework that generates virtually stained images using label-free tissue images, in which different stains are merged following a micro-structure map defined by the user. This approach uses a single deep neural network that receives two different sources of information as its input: (1) autofluorescence images of the label-free tissue sample and (2) a "digital staining matrix", which represents the desired microscopic map of the different stains to be virtually generated in the same tissue section. This digital staining matrix is also used to virtually blend existing stains, digitally synthesizing new histological stains. We trained and blindly tested this virtual-staining network using unlabelled kidney tissue sections to generate micro-structured combinations of haematoxylin and eosin (H&E), Jones' silver stain, and Masson's trichrome stain. Using a single network, this approach multiplexes the virtual staining of label-free tissue images with multiple types of stains and paves the way for synthesizing new digital histological stains that can be created in the same tissue cross section, which is currently not feasible with standard histochemical staining methods.Neural networks: Digital stains enhance pathologists' arsenal A machine learning approach uses combinations of multiple digital stains to highlight microscopic elements in a tissue sample, avoiding the delays, inconsistences and sometimes the need for multiple biopsies, all characteristic of traditional manual tissue staining techniques. The deep learning-based tissue staining framework was developed by Aydogan Ozcan, Yair Rivenson and colleagues at the University of California, Los Angeles. They used their approach to train neural networks to virtually stain kidney samples with one of three different types of stains or their combinations. Comparisons with manually stained tissue samples demonstrated that the virtual staining was highly accurate. The ability to use multiple digital stains and virtual stain blending on the same tissue sample using a single neural network could allow pathologists to get more relevant information from tissue and thus improve diagnoses.