直肠癌扩散加权成像b值选取及其对直肠癌显示能力的评价

Diffusion-weighted MR imaging of rectal cancer:determination of b values and evaluation of displaying ability

目的探讨扩散成像对直肠癌的显示能力和b值选取方法,确定直肠癌扩散成像的较佳b值。方法对32例原发性直肠癌患者术前应用1.5T磁共振行MRI常规和多b值扩散成像。计算肿瘤区域组织、正常肠壁、膀胱和背景噪声的信号强度值,计算不同图像中ADC值、信噪比及对比噪声比等数据,并进行统计学分析。结果扩散图像中肿瘤与正常肠壁的对比噪声比(CNR)随着b值的增加而下降,b值300、500、1000s/mm2图像CNR高于T2图像和b值1300s/mm2图像,有统计学意义(F=42.492,P<0.05)。扩散图像中膀胱SIR值均低于高分辨T2图像中膀胱SIR值,有统计学意义(F=144.786,P<0.05),b值升至800s/mm2时,SIR降低至2.437±0.844。随着b值的增高,肿瘤ADC值逐渐变小,有统计学意义(F=141.425,P<0.01)。结论b=1000s/mm2是直肠癌扩散成像的较佳b值,可以较好的克服灌注、T2穿透效应对直肠癌扩散成像的影响,准确反映组织的扩散特性。

Objective To investigate the ability of diffusiowweighted imaging （DWI） in displaying rectal cancer and determine reasonable b values of DWI of rectal cancer. Methods Thirty-two patients with rectal cancer with single-shot echo planar imaging （EPI） diffusion-weighted sequences （b=300, 500, 800, 1000 and 1300 s/mm^2） at 1.5T MR scanner with an 8 channel body phase-array coil. The imaging parameters were： TE/TR minimum/6000 ms, FOV 36 cm, Matrix 128×128, Slice thickness 5 ram, Nex 8. ASSET technology was used to reduce the distortion of EPI. The mean apparent diffusion coefficient （ADC） values and signal-to-noise ratio （SNR） of tumors, contrast-to-noise ratio （CNR） of tumors and normal rectal wall, signal intensity ratio （SIR） o{ bladder were calculated and compared. Results The CNR of tumors and normal rectal wall decreases following the descending of b values, and the CNR in the DWI with b= 300, 500, 1000 s/ram2 was significantly higher than that in DWI with b=1300 s/mm2 and T2-weighted imaging （F= 42. 492, P〈0.05）. The SIR of bladder in DWI （b=300, 500, 800, 1000 and 1300 s/mm^2） were lower than that in T2-weighted imaging （F=144. 786, P〈0.05）, when b values reached 800 s/mm^2 the SIR declined to 2. 437±0. 844. The mean ADC values of tumor were decreases following the ascending of b values. Conclusion b=1000 s/mm^2 is reasonable for DWI of rectal cancer, and that can get rid of the influence produced by blood perfusion and T2 shine through.