使用深度卷积生成对抗网络的高分辨率乳腺MRI重建

High-Resolution Breast MRI Reconstruction Using a Deep Convolutional Generative Adversarial Network.

作者信息

Sun Kun, Qu Liangqiong, Lian Chunfeng, Pan Yongsheng, Hu Dan, Xia Bingqing, Li Xinyue, Chai Weimin, Yan Fuhua, Shen Dinggang

机构信息

Department of Radiology, Ruijin Hospital, Shanghai Jiaotong University School of Medicine, Shanghai, China.

Department of Radiology and BRIC, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, USA.

出版信息

J Magn Reson Imaging. 2020 Dec;52(6):1852-1858. doi: 10.1002/jmri.27256. Epub 2020 Jul 12.

DOI:10.1002/jmri.27256

PMID:32656955

Abstract

BACKGROUND

A generative adversarial network could be used for high-resolution (HR) medical image synthesis with reduced scan time.

PURPOSE

To evaluate the potential of using a deep convolutional generative adversarial network (DCGAN) for generating HR and HR images based on their corresponding low-resolution (LR) images (LR and LR ).

STUDY TYPE

This was a retrospective analysis of a prospectively acquired cohort.

POPULATION

In all, 224 subjects were randomly divided into 200 training subjects and an independent 24 subjects testing set.

FIELD STRENGTH/SEQUENCE: Dynamic contrast-enhanced (DCE) MRI with a 1.5T scanner.

ASSESSMENT

Three breast radiologists independently ranked the image datasets, using the DCE images as the ground truth, and reviewed the image quality of both the original LR images and the generated HR images. The BI-RADS category and conspicuity of lesions were also ranked. The inter/intracorrelation coefficients (ICCs) of mean image quality scores, lesion conspicuity scores, and Breast Imaging Reporting and Data System (BI-RADS) categories were calculated between the three readers.

STATISTICAL TEST

Wilcoxon signed-rank tests evaluated differences among the multireader ranking scores.

RESULTS

The mean overall image quality scores of the generated HR and HR were significantly higher than those of the original LR and LR (4.77 ± 0.41 vs. 3.27 ± 0.43 and 4.72 ± 0.44 vs. 3.23 ± 0.43, P < 0.0001, respectively, in the multireader study). The mean lesion conspicuity scores of the generated HR and HR were significantly higher than those of the original LR and LR (4.18 ± 0.70 vs. 3.49 ± 0.58 and 4.35 ± 0.59 vs. 3.48 ± 0.61, P < 0.001, respectively, in the multireader study). The ICCs of the image quality scores, lesion conspicuity scores, and BI-RADS categories had good agreements among the three readers (all ICCs >0.75).

DATA CONCLUSION

DCGAN was capable of generating HR of the breast from fast pre- and postcontrast LR and achieved superior quantitative and qualitative performance in a multireader study.

LEVEL OF EVIDENCE

3 TECHNICAL EFFICACY STAGE: 2 J. MAGN. RESON. IMAGING 2020;52:1852-1858.

摘要

背景

生成对抗网络可用于减少扫描时间的高分辨率（HR）医学图像合成。

目的

评估使用深度卷积生成对抗网络（DCGAN）根据相应的低分辨率（LR）图像（LR和LR）生成HR和HR图像的潜力。

研究类型

这是一项对前瞻性采集队列的回顾性分析。

研究对象

总共224名受试者被随机分为200名训练受试者和一个独立的24名受试者测试集。

场强/序列：使用1.5T扫描仪进行动态对比增强（DCE）MRI。

评估

三名乳腺放射科医生以DCE图像作为参考标准，独立对图像数据集进行排名，并评估原始LR图像和生成的HR图像的质量。还对病变的BI-RADS类别和清晰度进行了排名。计算了三位读者之间平均图像质量评分、病变清晰度评分和乳腺影像报告和数据系统（BI-RADS）类别的组间/组内相关系数（ICC）。

统计检验

Wilcoxon符号秩检验评估多读者排名分数之间的差异。

结果

在多读者研究中，生成的HR和HR的平均整体图像质量评分显著高于原始LR和LR（分别为4.77±0.41对3.27±0.43和4.72±0.44对3.23±0.43，P<0.0001）。生成的HR和HR的平均病变清晰度评分显著高于原始LR和LR（分别为4.18±0.70对3.49±0.58和4.35±0.59对3.48±0.61，P<0.001）。图像质量评分、病变清晰度评分和BI-RADS类别的ICC在三位读者之间具有良好的一致性（所有ICC>0.75）。