深度学习重建算法在肠系膜上动脉CT血管成像中的应用评估

Evaluation of the application of deep learning reconstruction algorithm in superior mesenteric artery CT angiograpy

目的:评估深度学习图像重建算法在肠系膜上动脉CT血管成像(CTA)图像中的应用价值。方法:回顾性纳入40例患者肠系膜上动脉CTA的原始数据,利用滤波反投影重建(FBP)、自适应统计迭代重建(ASIR)30%、ASIR 60%、深度学习图像重建(DLIR)-L、DLIR-M、DLIR-H共6种算法重建图像,测量并分析肠系膜上动脉的CT值、SD值和信噪比,两位放射科医生进行主观图像评价。结果:6种算法重建图像CT值、噪声、信噪比和主观图像质量评分间差异有统计学意义(P<0.001),DLIR-H,DLIR-M、ASIR60%、DLIR-L、ASIR30%、FBP图像噪声依次增加,信噪比依次降低;3种DLIR的CT值无明显差异,但DLIR的CT值高于FBP和ASIR;主观评分显示DLIR重建图像血管锐利度增强。结论:DLIR改善了肠系膜上动脉CTA的图像质量,并且能提高血管锐利度,但也存在改变原始信息的风险。

Objective:To evaluate the application value of deep learning-based reconstruction(DLIR)algorithm in superior mesenteric artery CT angiography.Methods:The 40 patients'scan raw data of CT angiography of superior mesenteric artery were retrospectively included in this study.The images were reconstructed using six algorithms,including filtered back projection(FBP),adaptive statistical iterative reconstruction(ASIR)30%and 60%,deep learning image reconstruction(DLIR)-L,DLIR-M and DLIR-H.The CT value,SD value and signal to noise ratio of superior mesenteric artery were analyzed and two radiologists evaluated the subjective image quality.Results:There was a statistical difference(P<0.001)between the six algorithms in terms of CT value,noise,signal-to-noise ratio(SNR),and subjective image quality score for reconstructed images.Specifically,DLIR-H,DLIR-M,ASIR60%,DLIR-L,ASIR30%,and FBP images showed an increase in noise and a decrease in signal-to-noise ratio.There was no significant difference in CT values among the three types of DLIR,but the CT values of DLIR were higher than FBP and ASIR.Subjective scoring showed enhanced vascular sharpness in DLIR reconstructed images.Conclusion:DLIR can improve the image quality of CT angiography of superior mesenteric artery and improve the vessel sharpness,but there is also a risk of changing the original information.