Department of Imaging and Interventional Radiology, Faculty of Medicine, The Chinese University of Hong Kong, Hong Kong, Hong Kong SAR.
Department of Chemical Pathology, State Key Laboratory of Translational Oncology, The Chinese University of Hong Kong, Hong Kong, Hong Kong SAR.
Eur Radiol. 2021 Jun;31(6):3856-3863. doi: 10.1007/s00330-020-07451-y. Epub 2020 Nov 25.
A convolutional neural network (CNN) was adapted to automatically detect early-stage nasopharyngeal carcinoma (NPC) and discriminate it from benign hyperplasia on a non-contrast-enhanced MRI sequence for potential use in NPC screening programs.
We retrospectively analyzed 412 patients who underwent T2-weighted MRI, 203 of whom had biopsy-proven primary NPC confined to the nasopharynx (stage T1) and 209 had benign hyperplasia without NPC. Thirteen patients were sampled randomly to monitor the training process. We applied the Residual Attention Network architecture, adapted for three-dimensional MR images, and incorporated a slice-attention mechanism, to produce a CNN score of 0-1 for NPC probability. Threefold cross-validation was performed in 399 patients. CNN scores between the NPC and benign hyperplasia groups were compared using Student's t test. Receiver operating characteristic with the area under the curve (AUC) was performed to identify the optimal CNN score threshold.
In each fold, significant differences were observed in the CNN scores between the NPC and benign hyperplasia groups (p < .01). The AUCs ranged from 0.95 to 0.97 with no significant differences between the folds (p = .35 to .92). The combined AUC from all three folds (n = 399) was 0.96, with an optimal CNN score threshold of > 0.71, producing a sensitivity, specificity, and accuracy of 92.4%, 90.6%, and 91.5%, respectively, for NPC detection.
Our CNN method applied to T2-weighted MRI could discriminate between malignant and benign tissues in the nasopharynx, suggesting that it as a promising approach for the automated detection of early-stage NPC.
• The convolutional neural network (CNN)-based algorithm could automatically discriminate between malignant and benign diseases using T2-weighted fat-suppressed MR images. • The CNN-based algorithm had an accuracy of 91.5% with an area under the receiver operator characteristic curve of 0.96 for discriminating early-stage T1 nasopharyngeal carcinoma from benign hyperplasia. • The CNN-based algorithm had a sensitivity of 92.4% and specificity of 90.6% for detecting early-stage nasopharyngeal carcinoma.
为了将卷积神经网络（CNN）应用于非增强 MRI 序列，自动检测早期鼻咽癌（NPC）并将其与良性增生区分开来，从而为 NPC 筛查项目提供潜在用途。
我们回顾性分析了 412 例接受 T2 加权 MRI 检查的患者，其中 203 例经活检证实为局限于鼻咽部的原发性 NPC（T1 期），209 例为无 NPC 的良性增生。我们随机抽取了 13 例患者进行监测，以监控训练过程。我们应用了适用于三维 MR 图像的残差注意网络结构，并结合了切片注意机制，为 NPC 概率生成 0-1 的 CNN 评分。在 399 例患者中进行了三折交叉验证。使用学生 t 检验比较 NPC 组和良性增生组的 CNN 评分。采用接收者操作特征曲线下面积（AUC）来确定最佳 CNN 评分阈值。
在每一轮中，NPC 组和良性增生组的 CNN 评分差异均有统计学意义（p<0.01）。AUC 范围为 0.95-0.97，折间差异无统计学意义（p=0.35-0.92）。三折合并 AUC（n=399）为 0.96，最佳 CNN 评分阈值>0.71，对 NPC 检测的敏感性、特异性和准确性分别为 92.4%、90.6%和 91.5%。
本研究应用于 T2 加权 MRI 的 CNN 方法可以区分鼻咽部的恶性和良性组织，提示其作为自动检测早期 NPC 的一种有前途的方法。
• 基于卷积神经网络（CNN）的算法可以使用 T2 加权脂肪抑制 MR 图像自动区分恶性和良性疾病。• 基于 CNN 的算法对 T1 期早期 NPC 与良性增生的鉴别诊断准确率为 91.5%，受试者工作特征曲线下面积为 0.96。• 基于 CNN 的算法对早期 NPC 的检测敏感性为 92.4%，特异性为 90.6%。
