Abstract：
OBJECTIVE: The aim of this study was to investigate the imaging performance and quality differences of PCD-CT in standard resolution mode (SR) versus ultra-high resolution mode (UHR) in the lower extremity runoff of dose-matched CTAs in a human cadaveric model.
METHODS: Extracorporeal perfusion of the upper leg was established in one fresh-frozen human cadaver via inguinal and popliteal accesses using a peristaltic pump. Seven peripheral stents were deployed in the SFA. Photon-counting CTAs were performed under contrast perfusion in SR and UHR mode with dose-equivalent 120kVp acquisition protocols (low-/ medium-/ high-dose: CTDIVol=3, 5, 10 mGy) and reconstructed with four vascular convolution kernels. Lumen visibility and contrast-to-noise ratio were compared using analyses of variance. Subjective image quality was assessed using a pairwise, forced-choice comparison software.
RESULTS: Lumen visibility was equal for SR and UHR at the used dose levels. CNR increase by UHR was significant for (ultra-)sharp convolution kernels BV60 (3 mGy; UHR vs. SR, 19.9 ± 1.9 vs. 15.7 ± 1.6, p < 0.046) and BV76 (8.0 ± 0.6 vs. 5.4 ± 0.3, p < 0.001). The relative CNR increase was higher for low-dose than high-dose scans (BV76: 48% vs. 36% at high dose, p < 0.033). The CNR of the low-dose scan in UHR mode was comparable to the high-dose scan in SR mode when the ultra-sharp kernel was used (8.0 ± 0.6 vs. 9.1 ± 1.1, p > 0.760). Among UHR examinations, a significant increase in CNR could only be measured in BV76 (8.0 ± 0.6 (3 mGy) vs. 12.4 ± 0.9 (10 mGy), p < 0.001). Readers preferred subjective image quality of UHR for all kernels with BV76 being ranked highest.
CONCLUSIONS: The CNR increase in UHR mode is highest when combining low radiation dose and ultra-sharp reconstructions. Meanwhile, the subjective image quality in UHR mode generally supersedes SR images, suggesting further dose reduction potential.
METHODS: Extracorporeal perfusion of the upper leg was established in one fresh-frozen human cadaver via inguinal and popliteal accesses using a peristaltic pump. Seven peripheral stents were deployed in the SFA. Photon-counting CTAs were performed under contrast perfusion in SR and UHR mode with dose-equivalent 120kVp acquisition protocols (low-/ medium-/ high-dose: CTDIVol=3, 5, 10 mGy) and reconstructed with four vascular convolution kernels. Lumen visibility and contrast-to-noise ratio were compared using analyses of variance. Subjective image quality was assessed using a pairwise, forced-choice comparison software.
RESULTS: Lumen visibility was equal for SR and UHR at the used dose levels. CNR increase by UHR was significant for (ultra-)sharp convolution kernels BV60 (3 mGy; UHR vs. SR, 19.9 ± 1.9 vs. 15.7 ± 1.6, p < 0.046) and BV76 (8.0 ± 0.6 vs. 5.4 ± 0.3, p < 0.001). The relative CNR increase was higher for low-dose than high-dose scans (BV76: 48% vs. 36% at high dose, p < 0.033). The CNR of the low-dose scan in UHR mode was comparable to the high-dose scan in SR mode when the ultra-sharp kernel was used (8.0 ± 0.6 vs. 9.1 ± 1.1, p > 0.760). Among UHR examinations, a significant increase in CNR could only be measured in BV76 (8.0 ± 0.6 (3 mGy) vs. 12.4 ± 0.9 (10 mGy), p < 0.001). Readers preferred subjective image quality of UHR for all kernels with BV76 being ranked highest.
CONCLUSIONS: The CNR increase in UHR mode is highest when combining low radiation dose and ultra-sharp reconstructions. Meanwhile, the subjective image quality in UHR mode generally supersedes SR images, suggesting further dose reduction potential.
摘要：
目的：这项研究的目的是研究标准分辨率模式（SR）和超高分辨率模式（UHR）下PCD-CT在剂量下肢径流中的成像性能和质量差异匹配的CTA在人类尸体模型中。
方法：使用蠕动泵通过腹股沟和pop入路在一个新鲜冷冻的人尸体中建立了上腿的体外灌注。在SFA中部署了七个外周支架。光子计数CTA在SR和UHR模式下使用剂量等效120kVp采集协议（低/中/高剂量：CTDIVol=3、5、10mGy）在造影剂灌注下进行，并用四个血管卷积核进行重建。使用方差分析比较了管腔能见度和对比度噪声比。使用成对评估主观图像质量，强制选择比较软件。
结果：在使用的剂量水平下，SR和UHR的管腔能见度相等。对于（超）尖锐卷积核BV60（3mGy；UHRvs.SR,19.9±1.9vs.15.7±1.6，p<0.046）和BV76（8.0±0.6vs.5.4±0.3，p<0.001)。低剂量扫描的相对CNR增加高于高剂量扫描(BV76：48%vs.高剂量时36%，p<0.033)。当使用超尖锐内核时，UHR模式下低剂量扫描的CNR与SR模式下的高剂量扫描相当（8.0±0.6与9.1±1.1，p>0.760）。在UHR考试中，仅在BV76中可以测量到CNR的显着增加（8.0±0.6（3mGy）与12.4±0.9（10mGy），p<0.001)。读者更喜欢所有内核的UHR主观图像质量，其中BV76排名最高。
结论：当结合低辐射剂量和超尖锐重建时，UHR模式的CNR增加最高。同时,UHR模式下的主观图像质量通常取代SR图像，提示进一步的剂量减少潜力。
方法：使用蠕动泵通过腹股沟和pop入路在一个新鲜冷冻的人尸体中建立了上腿的体外灌注。在SFA中部署了七个外周支架。光子计数CTA在SR和UHR模式下使用剂量等效120kVp采集协议（低/中/高剂量：CTDIVol=3、5、10mGy）在造影剂灌注下进行，并用四个血管卷积核进行重建。使用方差分析比较了管腔能见度和对比度噪声比。使用成对评估主观图像质量，强制选择比较软件。
结果：在使用的剂量水平下，SR和UHR的管腔能见度相等。对于（超）尖锐卷积核BV60（3mGy；UHRvs.SR,19.9±1.9vs.15.7±1.6，p<0.046）和BV76（8.0±0.6vs.5.4±0.3，p<0.001)。低剂量扫描的相对CNR增加高于高剂量扫描(BV76：48%vs.高剂量时36%，p<0.033)。当使用超尖锐内核时，UHR模式下低剂量扫描的CNR与SR模式下的高剂量扫描相当（8.0±0.6与9.1±1.1，p>0.760）。在UHR考试中，仅在BV76中可以测量到CNR的显着增加（8.0±0.6（3mGy）与12.4±0.9（10mGy），p<0.001)。读者更喜欢所有内核的UHR主观图像质量，其中BV76排名最高。
结论：当结合低辐射剂量和超尖锐重建时，UHR模式的CNR增加最高。同时,UHR模式下的主观图像质量通常取代SR图像，提示进一步的剂量减少潜力。
