Generative Adversarial Networks for the Creation of Realistic Artificial Brain Magnetic Resonance Images
Generative Adversarial Networks for the Creation of Realistic Artificial Brain Magnetic Resonance Images
复制标题
DOI：
10.18383/j.tom.2018.00042
复制
发表时间：
2018-12-01
期刊：
影响因子：
1.9
通讯作者：
Rowe, Steven P.
中科院分区：
文献类型：
--
作者：
Kazuhiro, Koshino;Werner, Rudolf A.;Rowe, Steven P.
即使医疗数据集变得更加公开，大多数也仅限于特定的医疗条件。因此，机器学习方法的数据收集仍然具有挑战性，而合成数据增强，如生成对抗网络（GAN），可以克服这一障碍。在本质量控制研究中，由盲态放射科医生验证了基于深度卷积GAN（DCGAN）的人脑磁共振（MR）图像。共包括30名健康人和33名脑血管意外患者的96张T1加权脑图像。从T1加权图像生成训练数据集，并应用DCGAN生成额外的人工脑图像。由5名放射科医生（2名神经放射科医生[NR] vs 3名非神经放射科医生[NNR]）以二进制方式评价DCGAN创建与获取图像的可能性，以识别真实的与创建的图像。从数据集中随机选择图像（创建图像的变化，40%-60%）。所有研究图像均未被评定为未知。在创建的图像中，NR将45%和71%评定为真实的磁共振成像图像（NNR，24%、40%和44%）。相比之下，44%和70%的真实的图像被NR（NNR，10%、17%和27%）评定为生成图像。NR的准确度为0.55和0.30（NNR为0.83、0.72和0.64）。DCGAN创建的脑MR图像与采集的MR图像足够相似，因此在某些情况下无法区分。这样的人工智能算法可以有助于在各种临床应用中用于“数据饥饿”技术（诸如监督机器学习方法）的合成数据增强。
Even as medical data sets become more publicly accessible, most are restricted to specific medical conditions. Thus, data collection for machine learning approaches remains challenging, and synthetic data augmentation, such as generative adversarial networks (GAN), may overcome this hurdle. In the present quality control study, deep convolutional GAN (DCGAN)-based human brain magnetic resonance (MR) images were validated by blinded radiologists. In total, 96 T1-weighted brain images from 30 healthy individuals and 33 patients with cerebrovascular accident were included. A training data set was generated from the T1-weighted images and DCGAN was applied to generate additional artificial brain images. The likelihood that images were DCGAN-created versus acquired was evaluated by 5 radiologists (2 neuroradiologists [NRs], vs 3 non-neuroradiologists [NNRs]) in a binary fashion to identify real vs created images. Images were selected randomly from the data set (variation of created images, 40%-60%). None of the investigated images was rated as unknown. Of the created images, the NRs rated 45% and 71% as real magnetic resonance imaging images (NNRs, 24%, 40%, and 44%). In contradistinction, 44% and 70% of the real images were rated as generated images by NRs (NNRs, 10%, 17%, and 27%). The accuracy for the NRs was 0.55 and 0.30 (NNRs, 0.83, 0.72, and 0.64). DCGAN-created brain MR images are similar enough to acquired MR images so as to be indistinguishable in some cases. Such an artificial intelligence algorithm may contribute to synthetic data augmentation for "data-hungry" technologies, such as supervised machine learning approaches, in various clinical applications.