Abstract：
To determine if hepatic and splenic perfusion parameters are useful in identifying severe portal hypertension (SPH).
The study enrolled 52 patients who underwent perfusion CT scan within one week before the hepatic venous pressure gradient (HVPG) measurement. A commercial software package was used for post-processing to generate hepatic and splenic perfusion parameters. Correlations were assessed using Pearson and Spearman rank correlation coefficients. Logistic regression was used to screen predictive parameters of SPH. The cut-off values of parameters for severe portal hypertension were calculated, as well as the sensitivity and specificity.
There was a significant difference between SPH and non-severe portal hypertension (NSPH) in blood volume of liver (BVLiver), hepatic arterial fraction (HAF), hepatic arterial perfusion (HAP), portal venous perfusion (PVP), mean slope of increase in spleen (MSISpleen), BVSpleen, blood flow of spleen (BFSpleen), BVSpleen/Liver, and BVSpleen/Liver(P) (p < 0.05). The Spearman correlation coefficient was - 0.541 (p < 0.001) between BVSpleen/Live and HVPG and - 0.568 (p < 0.001) between BVSpleen/Liver(P) and HVPG. Using a BVSpleen/Liver value of 0.780 or BVSpleen/Liver(P) value of 1.061 as the cut-off value for the detection of SPH, the sensitivity and specificity were 94.7% and 72.7%, 100%, and 63.6% respectively.
There was a moderate correlation between CT perfusion parameters BVSpleen/Liver, BVSpleen/Liver(P), and HVPG, which may be used to detect severe portal hypertension.
The study enrolled 52 patients who underwent perfusion CT scan within one week before the hepatic venous pressure gradient (HVPG) measurement. A commercial software package was used for post-processing to generate hepatic and splenic perfusion parameters. Correlations were assessed using Pearson and Spearman rank correlation coefficients. Logistic regression was used to screen predictive parameters of SPH. The cut-off values of parameters for severe portal hypertension were calculated, as well as the sensitivity and specificity.
There was a significant difference between SPH and non-severe portal hypertension (NSPH) in blood volume of liver (BVLiver), hepatic arterial fraction (HAF), hepatic arterial perfusion (HAP), portal venous perfusion (PVP), mean slope of increase in spleen (MSISpleen), BVSpleen, blood flow of spleen (BFSpleen), BVSpleen/Liver, and BVSpleen/Liver(P) (p < 0.05). The Spearman correlation coefficient was - 0.541 (p < 0.001) between BVSpleen/Live and HVPG and - 0.568 (p < 0.001) between BVSpleen/Liver(P) and HVPG. Using a BVSpleen/Liver value of 0.780 or BVSpleen/Liver(P) value of 1.061 as the cut-off value for the detection of SPH, the sensitivity and specificity were 94.7% and 72.7%, 100%, and 63.6% respectively.
There was a moderate correlation between CT perfusion parameters BVSpleen/Liver, BVSpleen/Liver(P), and HVPG, which may be used to detect severe portal hypertension.
摘要：
目的：确定肝和脾灌注参数是否可用于识别严重门脉高压（SPH）。
方法：该研究纳入了52例患者，这些患者在肝静脉压力梯度（HVPG）测量前一周内接受了灌注CT扫描。商业软件包用于后处理以生成肝和脾灌注参数。使用Pearson和Spearman等级相关系数评估相关性。采用Logistic回归筛选SPH的预测参数。计算重度门静脉高压症的参数截断值，以及敏感性和特异性。
结果：SPH与非重度门静脉高压症（NSPH）的肝脏血容量（BVLiver）之间存在显着差异，肝动脉分数（HAF），肝动脉灌注（HAP），门静脉灌注（PVP），脾脏增加的平均斜率（MSIS脾脏），BVSplen,脾血流量（BFSplen），BV脾/肝，和BVSween/肝脏（P）（p<0.05）。Spearman相关系数在BVSween/Live和HVPG之间为-0.541（p<0.001），在BVSween/Liver（P）和HVPG之间为-0.568（p<0.001）。使用BVSween/Liver值为0.780或BVSween/Liver（P）值为1.061作为检测SPH的临界值，敏感性和特异性分别为94.7%和72.7%，100%,和63.6%。
结论：CT灌注参数BVSween/Liver之间存在中度相关性，BV脾/肝（P），和HVPG,可用于检测严重门脉高压。
方法：该研究纳入了52例患者，这些患者在肝静脉压力梯度（HVPG）测量前一周内接受了灌注CT扫描。商业软件包用于后处理以生成肝和脾灌注参数。使用Pearson和Spearman等级相关系数评估相关性。采用Logistic回归筛选SPH的预测参数。计算重度门静脉高压症的参数截断值，以及敏感性和特异性。
结果：SPH与非重度门静脉高压症（NSPH）的肝脏血容量（BVLiver）之间存在显着差异，肝动脉分数（HAF），肝动脉灌注（HAP），门静脉灌注（PVP），脾脏增加的平均斜率（MSIS脾脏），BVSplen,脾血流量（BFSplen），BV脾/肝，和BVSween/肝脏（P）（p<0.05）。Spearman相关系数在BVSween/Live和HVPG之间为-0.541（p<0.001），在BVSween/Liver（P）和HVPG之间为-0.568（p<0.001）。使用BVSween/Liver值为0.780或BVSween/Liver（P）值为1.061作为检测SPH的临界值，敏感性和特异性分别为94.7%和72.7%，100%,和63.6%。
结论：CT灌注参数BVSween/Liver之间存在中度相关性，BV脾/肝（P），和HVPG,可用于检测严重门脉高压。
