MRI动态增强扫描对前列腺癌的诊断价值

目的:在1.5T场强下评价MRI动态增强扫描诊断前列腺癌的能力。方法:对照经直肠前列腺穿刺活检,在24例患者的前列腺动态增强图像中测量130个ROI,其中癌肿组50个,良性组80个,建立时间-信号强度曲线,标注曲线的类型,在曲线上测得开始时间、达峰时间、初始值、峰值,计算时间差、差值、强化速率和强化比率,并进行统计学分析。结果:癌肿组50个ROI中,渐升型6个(12%),平台型5个(10%),速升缓降型39个(78%);良性组80个ROI中,渐升型20个(25%),平台型40个(50%),速升缓降型20个(25%)。癌肿组和良性组强化类型、开始时间、达峰时间、时间差、强化速率的差异有统计学意义(P<0.05),初始值、差值、峰值、强化比率的差异无统计学意义(P>0.05)。受试者工作特征曲线表明强化速率指标优于开始时间、达峰时间和时间差,其最佳工作点为72.085,特异度为0.76,敏感度为0.62,曲线下面积为0.73±0.046。结论:动态增强扫描是鉴别前列腺癌与良性组织的有效手段,并以强化速率指标最具诊断价值。

Value of MRI dynamic contrast-enhancement in diagnosis of prostate cancer

Objective： To evaluate MRI dynamic contrast-enhancement（DCE-MRI）in diagnosing prostate cancer with 1.5 T MR scanner. Methods：Based on the results of biopsy, we measured 130 regions of interest（RODin images of DCE MRI, comprised of 50 ROI＇s from cancer zone and 80 ROI＇s from non cancer zone. The time-signal intensity curve was made. The type of the curve, time of start, time to peak,initial value and peak value were recorded. Time difference,D-value, enhancement rate and ratio of enhanced were calculated. The data were analyzed with statistical methods. Results： 6 rising type ROI＇ s（ 12%）, 5 flat type ROI＇ s（10 %）, and 39 rapidly ascending and slowly descending type ROI＇ s （78 % ） were founded in cancer group. 20 rising type ROI＇s（25 %）,40 flat type ROI＇ s（50% ）, and 20 rapidly ascending and slowly descending type ROI＇s（25 % ）were founded in non-cancer group. There were significant differences between the malignant group and the benign group in type of enhancement, time of start, time to peak, time difference and enhancement rate（P〈0.05）. According to ROC curve, the enhancement rate was better than time of start,time to peak and time difference to distinguish prostate cancer from benign tissue. The optimal operating point（OOP） of ADC value was 72. 085, its specificity was 0.76, the sensitivity 0.62, the area under curve （AUC） 0.73 ± 0. 046. Conclusion：DCE-MRI is an effective method to differentiate prostate cancer from benign tissue. The enhancement rate is the factor with highest diagnosing value.