Shanghai Institute of Medical Imaging, 200032 Shanghai, China; Department of Interventional Radiology, Fudan University Zhongshan Hospital, 200032 Shanghai, China.
Shanghai Institute of Medical Imaging, 200032 Shanghai, China; Department of Interventional Radiology, Fudan University Zhongshan Hospital, 200032 Shanghai, China.
Diagn Interv Imaging. 2020 Feb;101(2):91-100. doi: 10.1016/j.diii.2019.07.002. Epub 2019 Jul 30.
The purpose of this study was to evaluate the ability of deep learning to differentiate pancreatic diseases on contrast-enhanced magnetic resonance (MR) images with the aid of generative adversarial network (GAN).
A total of 504 patients who underwent T1-weighted contrast-enhanced MR examinations before any treatments were included in this retrospective study. First, the MRI examinations of 398 patients (215 men, 183 women; mean age, 59.14±12.07 [SD] years [range: 16-85 years]) from one hospital were used as the training set. Then the MRI examinations of 50 (26 men, 24women; mean age, 58.58±13.64 [SD] years [range: 24-85 years]) and 56 (30 men, 26 women; mean age, 59.13±11.35 [SD] years [range: 26-80 years]) consecutive patients from two hospitals were separately collected as the internal and external validation sets. An InceptionV4 network was trained on the training set augmented by synthetic images from GANs. Classification performance of trained InceptionV4 network for every patch and every patient were made on both validation sets, respectively. The prediction agreement between convolutional neural network (CNN) and radiologist was measured by the Cohen's kappa coefficient.
The patch-level average accuracy and the micro-averaging area under receiver operating characteristic curve (AUC) of InceptionV4 network were 71.56% and 0.9204 (95% confidence interval [CI]: 0.9165-0.9308) for the internal validation set, and 79.46% and 0.9451 (95%CI: 0.9320-0.9523) for the external validation set, respectively. The patient-level average accuracy and the micro-averaging AUC of InceptionV4 network were 70.00% and 0.8250 (95%CI: 0.8147-0.8326) for the internal validation, 76.79% and 0.8646 (95%CI: 0.8489-0.8772) for the external validation set, respectively. Evaluated by human reader, the average accuracy and micro-averaging AUC for internal and external validation sets were 82.00% and 0.8950 (95%CI: 0.8817-0.9083), 83.93% and 0.9063 (95%CI: 0.8968-0.9212), respectively. The Cohen's kappa coefficients between InceptionV4 network and human reader for the internal and external invalidation sets were 0.8339 (95%CI: 0.6991-0.9447) and 0.8862 (95%CI: 0.7759-0.9738), respectively.
Deep learning using CNN and GAN had the potential to differentiate pancreatic diseases on contrast-enhanced MR images.
本研究旨在评估深度学习在基于生成对抗网络（GAN）的对比增强磁共振（MR）图像辅助下区分胰腺疾病的能力。
本回顾性研究纳入了 504 例在任何治疗前均行 T1 加权对比增强 MR 检查的患者。首先，使用来自一家医院的 398 例患者（215 名男性，183 名女性；平均年龄 59.14±12.07[标准差]岁[范围：16-85 岁]）的 MRI 检查作为训练集。然后，分别收集来自两家医院的 50 例（26 名男性，24 名女性；平均年龄 58.58±13.64[标准差]岁[范围：24-85 岁]）和 56 例（30 名男性，26 名女性；平均年龄 59.13±11.35[标准差]岁[范围：26-80 岁]）连续患者的 MRI 检查作为内部和外部验证集。基于训练集和 GAN 生成的合成图像，对 InceptionV4 网络进行训练。分别在两个验证集上对训练后的 InceptionV4 网络进行每个补丁和每个患者的分类性能评估。采用 Cohen's kappa 系数衡量 CNN 与放射科医生之间的预测一致性。
InceptionV4 网络在内部验证集的每个补丁和每个患者的平均准确率和微平均受试者工作特征曲线（AUC）下面积分别为 71.56%和 0.9204（95%置信区间[CI]：0.9165-0.9308），在外部验证集的平均准确率和微平均 AUC 分别为 79.46%和 0.9451（95%CI：0.9320-0.9523）。InceptionV4 网络在内部验证的每个患者和每个患者的平均准确率和微平均 AUC 分别为 70.00%和 0.8250（95%CI：0.8147-0.8326），在外部验证的平均准确率和微平均 AUC 分别为 76.79%和 0.8646（95%CI：0.8489-0.8772）。经人类读者评估，内部和外部验证集的平均准确率和微平均 AUC 分别为 82.00%和 0.8950（95%CI：0.8817-0.9083），83.93%和 0.9063（95%CI：0.8968-0.9212）。InceptionV4 网络与人类读者在内部和外部验证集的 Cohen's kappa 系数分别为 0.8339（95%CI：0.6991-0.9447）和 0.8862（95%CI：0.7759-0.9738）。
使用 CNN 和 GAN 的深度学习有潜力在对比增强 MR 图像上区分胰腺疾病。
