多期肝脏CT成像中的迭代重建:图像质量的定性与基于任务的分析
本研究对65例常规多期肝脏CT患者的图像分别用FBP、ASIR50、ASIR80和MBIR重建,定性评估肝动脉和门静脉最远端可见分支,定量测量肝实质、肝动脉和门静脉的对比噪声比,并与体模上计算的检出指数d'对比。结果显示MBIR和ASIR80较FBP能更远端显示肝动脉和门静脉(p≤0.001),d'与血管分支显示等级强相关(r=0.96和0.95),提示MBIR和ASIR可显著改善多期肝脏CT图像质量。
要点
- MBIR和ASIR80较FBP能更远端显示肝动脉和门静脉(p≤0.001)。
- 体模检出指数d'与肝动脉、门静脉分支显示等级强相关(r=0.96和0.95)。
- 患者与体模间CNR相关性较弱(肝动脉r=0.76,门静脉r=0.80)。
核心结论
- MBIR和ASIR可显著改善多期肝脏CT的图像质量,尤其表现为能更远端地检出肝动脉和门静脉,这与体模上估算的基于任务的检出指数d'一致。
研究设计与样本
- 对65例在1个月内接受常规多期肝脏CT的患者图像分别进行FBP、ASIR50、ASIR80和MBIR重建,并采用与患者相同CT协议扫描的体模进行对比分析。
关键影像表现/征象
- 定性评估显示MBIR和ASIR80较FBP能更远端显示肝动脉和门静脉(p≤0.001)。
- 定量测量肝实质、肝动脉和门静脉的对比噪声比(CNR),并计算体模的检出指数d'。
- 患者与体模间CNR相关性较弱(肝动脉r=0.76,门静脉r=0.80),而d'与肝动脉和门静脉的分支显示等级呈强相关(r=0.96和0.95)。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
英文原文(摘要)· 点击展开
AIM: To evaluate the clinical benefits on image quality (IQ) of adaptive statistical iterative reconstruction (ASIR) and model-based iterative reconstruction (MBIR) in multiphasic liver CT compared to filtered back-projection (FBP), in patients and on phantoms using a novel task-based metric.
MATERIALS AND METHODS: Image data of 65 patients who underwent a routine multiphasic liver CT during a 1-month period were reconstructed with FBP, ASIR50, ASIR80, and MBIR. IQ was assessed qualitatively by ranking the most distal hepatic artery (HA) and portal vein (PV) visible; and quantitatively by measuring contrast-to-noise ratio (CNR) of the liver parenchyma, HA and PV. IQ was compared between each reconstruction and correlated to CNR and detectability index (d') measurements computed on phantoms scanned with the same CT protocol as for patients.
RESULTS: HA and PV were seen more distally on MBIR and ASIR80 compared to FBP (p≤0.001). The CNR correlated weakly between patient and phantom (r=0.76 and 0.80 for HA and PV, respectively), whereas d' correlated strongly with the division order of HA and PV (r=0.96 and 0.95, respectively).
CONCLUSION: MBIR and ASIR significantly improve the IQ of multiphasic liver CT, especially through better distal detection of HA and PV, in agreement with the adapted task-based metric d' estimated on phantoms.
引用
- Pasquier H, Gardavaud F, Chiaradia M, et al. Iterative reconstructions in multiphasic CT imaging of the liver: qualitative and task-based analyses of image quality. Clin Radiol. 2018. PMID: 29929903. doi:10.1016/j.crad.2018.05.006
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。