评估结直肠癌256层螺旋CT全肿瘤灌注成像与单层面灌注成像一致性的初步研究

Assessment of consistency of the whole tumor and single section perfusion imaging with 256-slice spiral CT： a preliminary study

目的 探讨结直肠癌CT全肿瘤与肿瘤最大层面灌注成像灌注参数的一致性及测量的可重复性。方法 收集22例经病理证实的结直肠腺癌患者行256层螺旋CT灌注成像检查,获得全肿瘤灌注图像,2名医师分别测量记录每个肿瘤最大层面以及每个肿瘤所有层面各项灌注参数值的均值,其中1名医师间隔3个月后重复上述测量。测量的参数值包括血流量（BF）、血容量（BV）及达峰时间（TTP）。采用配对t检验比较2种分析方法所得灌注参数的差异,采用Bland-Altman检验评估相同观察者及不同观察者重复测量的一致性.结果 22例患者顺利完成检查.采用全肿瘤多层面方法得到的BF、BV及TTP分别为（35.59±14.59）ml·min^-1·100g^-1、（17.55±4.21）ml·100 g-1和（21.30 ±7.57）s,采用肿瘤最大层面方法得到的上述值分别为（34.64±13.29）ml·min^-1·100g^-1、（17.61 ±6.39）ml·100g^-1和（19.82 ±9.01）s,差异均无统计学意义（t值分别为0.218、-0.033和-0.668,P值均＞0.05）.同一医师前后2次采用全肿瘤多层面方法,所得BF、BV、TTP差异度的95％一致性范围分别为（-5.3 ～10.0）％、（-13.8～10.8）％和（-15.0～12.6）％,采用肿瘤最大层面方法,所得BF、BV、TTP结果95％一致性范围分别为（-14.3 ～ 16.5）％、（-24.2 ～ 22.2）％和（-19.0～16.1）％；2名医师采用全肿瘤多层面灌注分析方法,所得BF、BV、TTP差异度的95％一致性范围分别为（-8.0～8.3）％、（-10.9～11.5）％和（-14.5～11.1）％,采用肿瘤最大层面灌注分析方法,所得上述结果的95％一致性范围分别为（-10.2～14.1）％、（-19.0～17.6）％和（-22.0 ～24.0）％,说明采用全肿瘤灌注分析方法重复性更好.结论 肿瘤最大层面与全肿瘤多层面平均灌注分析方法所得灌注结果差异无统计学意义,但后者对灌注参数测量的可重复性更高,更能反映肿瘤整体的血管生成情况。

Objective To determine the consistency between quantitative CT perfusion measurements of colorectal cancer obtained from single section with maximal tumor dimension and from average of whole tumor,and compare intra-and inter-observer consistency of the two analysis methods.Methods Twenty-two patients with histologically proven colorectal cancer were examined prospectively with 256-slice CT and the whole tumor perfusion images were obtained.Perfusion parameters were obtained from region of interest （ROI） inserted in single section showing maximal tumor dimension,then from ROI inserted in all tumor-containing sections by two radiologists.Consistency between values of blood flow （BF）,blood volume （BV） and time to peak （TTP） calculated by two methods was assessed.Intra-observer consistency was evaluated by comparing repeated measurements done by the same radiologist using both methods after 3 months.Perfusion measurements were done by another radiologist independently to assess inter-observer consistency of both methods.The results from different methods were compared using paired t test and Bland-Altmnan plot.Results Twenty-two patients were examined successfully.The perfusion parameters BF,BV and TTP obtained by whole tumor perfusion and single-section analysis were （35.59 ±14.59） ml · min^-1 · 100 g^-1,（17.55±4.21） ml · 100 g^-1,（21.30 ±7.57） s and （34.64 ± 13.29）ml·min^-1·100 g^-1,（17.61±6.39）ml·100 g^-1,（19.82±9.01）s,respectively.No significant differences were observed between the means of the perfusion parameters （BF,BV,TTP）calculated by the two methods （t =0.218,-0.033,-0.668,P 〉 0.05,respectively）.The intra-observer 95% limits of consistency of perfusion parameters were BF-5.3% to 10.0%,BV-13.8% to 10.8%,TTP-15.0% to 12.6% with whole tumor analysis,respectively; BF-14.3% to 16.5%,BV-24.2% to 22.2%,TTP-19.0% to 16.1% with single section analysis,respectively.The inter-observer 95% limits of consistency of perfusion parameters were BF-8.0% to 8.3%,BV-10.9% to 11.5%,TTP -14.5% to 11.1% with whole volume analysis,respectively; BF-10.2% to 14.1%,BV-19.0% to 17.6%,TTP-22.0% to 24.0% with single section analysis,respectively.Conclusion There was no statistically different between the single section and whole volume analysis of tumor perfusion CT.The whole volume perfusion analysis apparently improves intra-and inter-observer consistency and can reflect the whole tumor angiogenesis more accurately and repeatedly.