节点文献

深度学习图像重建算法在低剂量腹部增强CT中的应用

The application of deep learning image reconstruction algorithm in low-dose contrast-enhanced abdominal CT

  • 推荐 CAJ下载
  • PDF下载
  • 不支持迅雷等下载工具,请取消加速工具后下载。

【作者】 王祎泽王铭君李露露高照吴兴旺

【Author】 WANG Yize;WANG Mingjun;LI Lulu;GAO Zhao;WU Xingwang;Department of Radiology,the First Affiliated Hospital of Anhui Medical University;CT Research Centre GE China;

【通讯作者】 吴兴旺;

【机构】 安徽医科大学第一附属医院放射科GE中国CT影像研究中心

【摘要】 目的 探讨深度学习图像重建算法(DLIR)对低剂量腹部增强CT图像质量的影响。方法 前瞻性选取52例行腹部增强CT扫描的患者,动脉期使用低剂量扫描模式,门静脉期及延迟期使用常规剂量扫描模式。动脉期图像使用40%自适应统计迭代重建算法(ASIR-V)、80%ASIR-V (层厚1.25 mm和层厚5 mm)(ASIR-V为前置)与中水平DLIT(DLIR-M)、高水平DLIT(DLIR-H)(层厚1.25 mm)重建;门静脉期图像使用40%ASIR-V(层厚1.25 mm)进行重建。测量肝脏、主动脉、脾脏、竖脊肌的CT值,将皮下脂肪的标准差(SD)值作为背景噪声。计算对比噪声比(CNR)作为图像客观评估的结果。图像的主观评价由2名放射科医师双盲法进行。结果 低剂量下同一层厚,与40%ASIR-V、80%ASIR-V、DLIR-M相比,DLIR-H有着最低的SD(8.63 HU±2.35 HU)、最高的CNR与主观评分(4.79分±0.41分)。低剂量下DLIR-H(层厚1.25 mm)的CNR低于80%ASIR-V(层厚5 mm),DLIR-M(层厚1.25 mm)的CNR低于40%ASIR-V(层厚5 mm),但是DLIR-H(层厚1.25 mm)的主观评分好于40%ASIR-V(层厚5 mm)。低剂量下DLIR-H的SD值(8.63 HU±2.35 HU)小于常规剂量下40%ASIR-V的SD值(11.54 HU±2.97 HU)。结论与ASIR-V相比,DLIR能够更好地降低图像噪声,提高图像质量,在低剂量腹部增强CT中具有更高的临床应用价值。

【Abstract】 Objective To explore the influence of deep learning image reconstruction(DLIR) algorithm on image quality of lowdose contrast-enhanced abdominal CT.Methods Fifty-two patients who would undergo contrast-enhanced abdominal CT were prospectively selected.Low-dose CT was used in the arterial phase and conventional-dose CT was used in the portal vein phase and delayed phase.Arterial phase images were reconstructed by using 40% adaptive statistical iterative reconstruction-V(ASIR-V),80% ASIR-V(1.25 mm and 5 mm thickness)(ASIR-V was prepositioned),medium level DLIR(DLIR-M) and high level DLIR(DLIR-H)(1.25 mm thickness).Portal vein phase images were reconstructed by using 40% ASIR-V(1.25 mm thickness).The CT values of liver,aorta,spleen and erector spinal muscle were measured,and standard deviation(SD) values of subcutaneous fat were measured as background noise.Contrast-to-noise ratio(CNR) was calculated as the result of image objective evaluation.Subjective evaluation of the images was performed by two radiologists in a double-blind method.Results Compared with 40% ASIR-V,80% ASIR-V and DLIR-M in the same sclice thickness at low dose.DLIR-H had the lowest background noise with SD of(8.63 HU±2.35 HU),the highest CNR and subjective score(4.79 branch±0.41 branch).At low dose,the CNR of DLIR-H(1.25 mm thickness)and DLIR-M(1.25 mm thickness) was lower than that of 80% or 40% ASIR-V(5 mm thickness),but the subjective score of DLIR-H(1.25 mm thickness) was greater than that of 40% ASIR-V(5 mm thickness).The SD values of DLIR-H at low dose(8.63 HU±2.35 HU)were lower than those of 40%ASIR-V at conventional dose(11.54 HU±2.97 HU).Conclusion Compared with ASIR-V,DLIR can significantly reduce image noise and improve image quality,and has higher clinical value in low-dose contrast-enhanced abdominal CT.

  • 【文献出处】 实用放射学杂志 ,Journal of Practical Radiology , 编辑部邮箱 ,2023年03期
  • 【分类号】TP18;TP391.41;R814
  • 【下载频次】18
节点文献中: 

本文链接的文献网络图示:

本文的引文网络