Successful Identification of Nasopharyngeal Carcinoma in Nasopharyngeal Biopsies Using Deep Learning
Successful Identification of Nasopharyngeal Carcinoma in Nasopharyngeal Biopsies Using Deep Learning
复制标题
DOI：
10.3390/cancers12020507
复制
发表时间：
2020-02-01
期刊：
影响因子：
5.2
通讯作者：
Yeh, Chao-Yuan
中科院分区：
文献类型：
--
作者：
Chuang, Wen-Yu;Chang, Shang-Hung;Yeh, Chao-Yuan
鼻咽癌（NPC）的病理诊断具有挑战性，因为大多数病例是非角化性癌，分化程度低，有大量淋巴细胞混合。我们的目的是评估使用深度学习在鼻咽活检中识别NPC的可能性。共纳入726例鼻咽活检。其中，随机抽取100例作为测试集，20例作为验证集，其余606例作为训练集。所有三个数据集的NPC病例和良性病例数量相等。进行手动注释。裁剪的256 x 256像素的正方形图像块用于块级训练、验证和测试。最终的斑块级算法有效地识别了NPC斑块，其受试者工作特征曲线（AUC）下的面积为0.9900。使用梯度加权类激活映射，我们证明了NPC斑块的识别是基于肿瘤细胞的形态学特征。在第二阶段，将整个载玻片图像依次裁剪成补丁，使用补丁级算法进行推断，并重建成较小尺寸的图像用于训练、验证和测试。最后，用于玻片水平鉴别NPC的AUC为0.9848。我们的研究结果首次表明，深度学习算法可以识别NPC。
Pathologic diagnosis of nasopharyngeal carcinoma (NPC) can be challenging since most cases are nonkeratinizing carcinoma with little differentiation and many admixed lymphocytes. Our aim was to evaluate the possibility to identify NPC in nasopharyngeal biopsies using deep learning. A total of 726 nasopharyngeal biopsies were included. Among them, 100 cases were randomly selected as the testing set, 20 cases as the validation set, and all other 606 cases as the training set. All three datasets had equal numbers of NPC cases and benign cases. Manual annotation was performed. Cropped square image patches of 256 x 256 pixels were used for patch-level training, validation, and testing. The final patch-level algorithm effectively identified NPC patches, with an area under the receiver operator characteristic curve (AUC) of 0.9900. Using gradient-weighted class activation mapping, we demonstrated that the identification of NPC patches was based on morphologic features of tumor cells. At the second stage, whole-slide images were sequentially cropped into patches, inferred with the patch-level algorithm, and reconstructed into images with a smaller size for training, validation, and testing. Finally, the AUC was 0.9848 for slide-level identification of NPC. Our result shows for the first time that deep learning algorithms can identify NPC.