Zhang Yicun, Wang Luotong, Yuan Dian, Qi Ke, Zhang Mengyuan, Zhang Weiting, Gao Jianbo, Liu Jie
Department of Radiology, The First Affiliated Hospital of Zhengzhou University, Zhengzhou, China.
CT Imaging Research Center, GE HealthCare China, Beijing, China.
Eur Radiol. 2025 Jul 2. doi: 10.1007/s00330-025-11764-1.
This study aims to assess the feasibility of "double-low," low radiation dosage and low contrast media dosage, CT pulmonary angiography (CTPA) based on deep-learning image reconstruction (DLIR) algorithms.
One hundred consecutive patients (41 females; average age 60.9 years, range 18-90) were prospectively scanned on multi-detector CT systems. Fifty patients in the conventional-dose group (CD group) underwent CTPA with 100 kV protocol using the traditional iterative reconstruction algorithm, and 50 patients in the low-dose group (LD group) underwent CTPA with a 70 kVp DLIR protocol. Radiation and contrast agent doses were recorded and compared between groups. Objective parameters were measured and compared. Two radiologists evaluated images for overall image quality, artifacts, and image contrast separately on a 5-point scale. The furthest visible branches were compared between groups.
Compared to the control group, the study group reduced the dose-length product by 80.3% (p < 0.01) and the contrast media dose by 33.3%. CT values, SD values, signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR) showed no statistically significant differences (all p > 0.05) between the LD and CD groups. The overall image quality scores were comparable between the LD and CD groups (p > 0.05), with good in-reader agreement (k = 0.75). More peripheral pulmonary vessels could be assessed in the LD group compared with the CD group.
70 kVp combined with DLIR reconstruction for CTPA can further reduce radiation and contrast agent dose while maintaining image quality and increasing the visibility on the pulmonary artery distal branches.
Question Elevated radiation exposure and substantial doses of contrast media during CT pulmonary angiography (CTPA) augment patient risks. Findings The "double-low" CT pulmonary angiography protocol can diminish radiation doses by 80.3% and minimize contrast doses by one-third while maintaining image quality. Clinical relevance With deep learning algorithms, we confirmed that CTPA images maintained excellent quality despite reduced radiation and contrast dosages, helping to reduce radiation exposure and kidney burden on patients. The "double-low" CTPA protocol, complemented by deep learning image reconstruction, prioritizes patient safety.
本研究旨在评估基于深度学习图像重建（DLIR）算法的“双低”（低辐射剂量和低对比剂剂量）CT肺血管造影（CTPA）的可行性。
对100例连续患者（41例女性；平均年龄60.9岁，范围18 - 90岁）进行多排CT系统前瞻性扫描。常规剂量组（CD组）50例患者采用传统迭代重建算法，以100 kV方案进行CTPA检查；低剂量组（LD组）50例患者采用70 kVp的DLIR方案进行CTPA检查。记录并比较两组的辐射剂量和对比剂剂量。测量并比较客观参数。两名放射科医生分别以5分制对图像的整体图像质量、伪影和图像对比度进行评估。比较两组间最远可见分支情况。
与对照组相比，研究组的剂量长度乘积降低了80.3%（p < 0.01），对比剂剂量降低了33.3%。LD组和CD组之间的CT值、标准差（SD）值、信噪比（SNR）和对比噪声比（CNR）均无统计学显著差异（所有p > 0.05）。LD组和CD组的整体图像质量评分相当（p > 0.05），阅片者间一致性良好（k = 0.75）。与CD组相比，LD组能够评估更多的外周肺血管。
70 kVp联合DLIR重建用于CTPA可进一步降低辐射剂量和对比剂剂量，同时保持图像质量并提高肺动脉远端分支的可视性。
问题 CT肺血管造影（CTPA）期间辐射暴露增加和大量对比剂剂量会增加患者风险。发现 “双低”CT肺血管造影方案可在保持图像质量的同时，将辐射剂量降低80.3%，并将对比剂剂量降至最低三分之一。临床意义 通过深度学习算法，我们证实尽管辐射剂量和对比剂剂量降低，但CTPA图像仍保持优异质量，有助于减少患者的辐射暴露和肾脏负担。辅以深度学习图像重建的“双低”CTPA方案将患者安全放在首位。
