医学
核医学
图像质量
图像噪声
成像体模
有效剂量(辐射)
辐射剂量
B组
对比噪声比
放射科
外科
计算机科学
图像(数学)
人工智能
作者
Jiang Shi,Xu Dong,Hong Zhi Dai,Li Shen,Yi ding Ji
摘要
Objective: To assess radiation dose and image quality of chest CT examinations in low-weight children acquired at ultralow tube voltage (70 kVp) combined with Flash scan technique. Materials and methods: 30 consecutive paediatric patients (weight <20 kg) required non-contrast chest CT at 70 kVp with Flash scan mode (Group A). 30 patients for paediatric standard 80-kVp protocols with conventional spiral mode (Group B) were selected from the picture archiving and communication system. For each examination, the volume CT dose index (CTDIvol) and dose–length product (DLP), and the effective dose (adapted as 16-cm phantom) (ED16cm) were estimated. The image noise, signal-to-noise ratio (SNR), overall subjective image quality and respiratory motion artefacts were evaluated. Results: For radiation dose, CTDIvol (mGy), DLP (mGy cm) and ED16cm (mSv) of Group A were significantly lower than those of Group B [CTDIvol: 0.48 ± 0.003 mGy (Group A) vs 0.80 ± 0.005 mGy (Group B); p<0.001 DLP: 10.23 ± 1.35 mGy cm (Group A) vs 15.6 ± 2.02 mGy cm (Group B); p<0.001 ED16cm: 0.61 ± 0.91 mSv (Group A) vs 0.89 ± 0.13 mSv (Group B); p<0.001]. The mean image noise with Group A increased 28.5% (p = 0.002), and the mean SNR decreased 14.8% compared with Group B (p = 0.193). There was no statistical difference in overall subjective image quality grades, and Group A had significantly lower respiratory motion artefact grades than Group B (p < 0.001). Conclusion: Ultralow tube voltage (70 kVp) combined with the Flash scan technique of the chest can obtain images with clinically acceptable image noise and minimum respiratory motion artefacts in low-weight children, whilst reducing radiation dose significantly. Advances in knowledge: The feasibility of chest CT scan in low-weight children with ultralow tube voltage (70 kVp) combined with Flash scan technique has firstly been evaluated in our study.
科研通智能强力驱动
Strongly Powered by AbleSci AI