基于多尺度残差去噪生成对抗网络的新型超分辨率算法提高计算机断层血管造影对颅内大动脉狭窄的诊断性能

Improving the diagnostic performance of computed tomography angiography for intracranial large arterial stenosis by a novel super-resolution algorithm based on multi-scale residual denoising generative adversarial network.

作者信息

Sun Jun, Li Zhang-Yu, Li Peng-Cheng, Li Hao, Pang Xiong-Wen, Wang Hui

机构信息

Department of Neurosurgery, the Third Affiliated Hospital of Sun Yat-sen University, Guangzhou 510630, China.

School of Computer Science, South China Normal University, Guangzhou 510631, China.

出版信息

Clin Imaging. 2023 Apr;96:1-8. doi: 10.1016/j.clinimag.2023.01.009. Epub 2023 Jan 27.

DOI:10.1016/j.clinimag.2023.01.009

PMID:36731372

Abstract

BACKGROUND

Computed tomography angiography (CTA) is very popular because it is characterized by rapidity and accessibility. However, CTA is inferior to digital subtraction angiography (DSA) in the diagnosis of intracranial artery stenosis or occlusion. DSA is an invasive examination, so we optimized the quality of cephalic CTA images.

METHODS

We used 5000 CTA images to train multi-scale residual denoising generative adversarial network (MRDGAN). And then 71 CTA images with intracranial large arterial stenosis were treated by Super-Resolution based on Generative Adversarial Network (SRGAN), Enhanced Super-Resolution based on Generative Adversarial Network (ESRGAN) and post-trained MRDGAN, respectively. Peak signal-to-noise ratio (PSNR) and structural similarity index measurement (SSIM) of the SRGAN, ESRGAN, MRDGAN and original CTA images were measured respectively. The qualities of MRDGAN and original images were visually assessed using a 4-point scale. The diagnostic coherence of digital subtraction angiography (DSA) with MRDGAN and original images was analyzed.

RESULTS

The PSNR was significantly higher in the MRDGAN CTA images (35.96 ± 1.51) than in the original (31.51 ± 1.43), SRGAN (25.75 ± 1.18) and ESRGAN (30.36 ± 1.05) CTA images (all P < 0.001). The SSIM was significantly higher in the MRDGAN CTA images (0.95 ± 0.02) than in the SRGAN (0.88 ± 0.03) and ESRGAN (0.90 ± 0.02) CTA images (all P < 0.01). The visual assessment was significantly higher in the MRDGAN CTA images (3.52 ± 0.58) than in the original CTA images (2.39 ± 0.69) (P < 0.05). The diagnostic coherence between MRDGAN and DSA (κ = 0.89) was superior to that between original images and DSA (κ = 0.62).

CONCLUSION

Our MRDGAN can effectively optimize original CTA images and improve its clinical diagnostic value for intracranial large artery stenosis.

摘要

背景

计算机断层血管造影（CTA）因其快速性和易获得性而非常受欢迎。然而，CTA在颅内动脉狭窄或闭塞的诊断方面不如数字减影血管造影（DSA）。DSA是一种侵入性检查，因此我们优化了头部CTA图像的质量。

方法

我们使用5000张CTA图像训练多尺度残差去噪生成对抗网络（MRDGAN）。然后，分别使用基于生成对抗网络的超分辨率（SRGAN）、基于生成对抗网络的增强超分辨率（ESRGAN）和训练后的MRDGAN对71张颅内大动脉狭窄的CTA图像进行处理。分别测量SRGAN、ESRGAN、MRDGAN和原始CTA图像的峰值信噪比（PSNR）和结构相似性指数测量值（SSIM）。使用4分制对MRDGAN和原始图像的质量进行视觉评估。分析数字减影血管造影（DSA）与MRDGAN和原始图像的诊断一致性。

结果

MRDGAN处理后的CTA图像的PSNR（35.96±1.51）显著高于原始图像（31.51±1.43）、SRGAN处理后的图像（25.75±1.18）和ESRGAN处理后的图像（30.36±1.05）（所有P<0.001）。MRDGAN处理后的CTA图像的SSIM（0.95±0.02）显著高于SRGAN处理后的图像（0.88±0.03）和ESRGAN处理后的图像（0.90±0.02）（所有P<0.01）。MRDGAN处理后的CTA图像的视觉评估得分（3.52±0.58）显著高于原始CTA图像（2.39±0.69）（P<0.05）。MRDGAN与DSA之间的诊断一致性（κ=0.89）优于原始图像与DSA之间的诊断一致性（κ=0.62）。