目的 探讨直肠癌CT灌注成像参 数特点及其与病理分化程度、临床分期 的关系。方法 选取24例直肠癌患者治 疗前行CT灌注成像扫描，绘制靶血管、 病灶及其邻近正常直肠组织感兴趣区 (region of interest，ROI)的时间-密 度曲线(time density curve，TDC)，记 录病灶、正常直肠组织的血流量(blood flow，BF)、血容量(blood volume， B V )、平均通过时间 ( m e a n t r a n s i t time，MTT)、血管表面通透性(vascular Permeability surfaceareaproduct， PS)和峰值CT值，并于术后对不同TNM分 期及病理分化程度的各灌注参数值进行 比较。结果 直肠癌的BF、BV、PS值、 峰值CT值明显高于正常组织，MTT值明 显低于正常组织，差异有统计学意义(t 值分别为-3.36、3.18、3.20、4.88、- 2.21，均P<0.05)。不同临床分期组间的 各灌注参数值差异均无统计学意义（t值 分别为-0.73、-0.86、-0.28、-1.28， 均P>0.05）。高中分化组直肠癌的BF值 明显高于低分化组，差异有统计学意义 (t=-3.35，P<0.05)。结论 多层螺旋CT 灌注成像能一定程度上反映肿瘤内部的 血流灌注情况，CT灌注参数对直肠癌与 正常组织鉴别、直肠癌分化程度的鉴别 均有一定参考意义。

Objective To evaluate the relationship between the characteristics of CT perfusion parameters and the degree of pathologic differentiation as well as TNM staging in rectal cancer. Methods 24 patents with histological verified rectal cancer underwent multi-slice spiral CT perfusion scanning before beginning therapy. Time density curves(TDC) were created from the region of interest (ROI) in the target vessels, lesions and normal rectal tissue adjacent to the lesion. The parameters of ROI were measured, such as blood flow(BF), blood volume(BV), mean transit time(MTT), vascular permeability surface area product(PS) and CT enhanced peak ,then determined TNM stages and the degree of tumor differentiation after surgery. The correlation between the parameters and the clinical and pathological factors above was studied. Results The rectal cancer tissue showed significantly higher BF, BV , PS and CT enhanced peak (P<0.05) and showed significantly shorter MTT (P<0.05) compared with those of normal tissue. BF in group of higher degree of differentiation was significantly higher than those in poorly differentiated tumor (P<0.05), the parameters were not associated with disease stage(P>0.05). Conclusion To a certain degree, Multi-slice spiral CT perfusion imaging can reflect the blood flow perfusion of the internal tumors, and CT perfusion parameters have a certain value for differentiation of rectal cancer and normal tissues as well as identification of different degree of pathological differentiation.