SCDU Radiologia, Università del Piemonte Orientale A. Avogadro, C.so Mazzini 18, 28100, Novara, Italy.
Radiol Med. 2011 Aug;116(5):720-33. doi: 10.1007/s11547-011-0629-1. Epub 2011 Feb 1.
We retrospectively evaluated the role of diffusion-weighted imaging (DWI) with fat and background signal suppression in the differential diagnosis of benign and malignant lesions of the chest-mediastinum by calculating the mean apparent diffusion coefficient (ADC) values.
Thirty-four patients with lung nodules/mediastinal masses underwent magnetic resonance (MR) imaging of the chest with conventional and DWI sequences. All patients had been previously studied with computed tomography (CT). After magnetic resonance (MR) imaging the patients underwent transthoracic CT-guided biopsy or mediastinoscopy. After the histopathological diagnosis had been obtained, the lesions were retrospectively divided into five groups: adenocarcinomas (n=16), squamous cell carcinomas (n=12), chronic pneumonias (n=2), malignant mediastinal tumours (n=2) and typical carcinoids (n=2). We compared ADC values in the different lesion groups using the Mann-Whitney U test.
There were statistically significant differences (p<0.05) between ADC values of benign and malignant lesions. Using an ADC value of 1.25×10⁻³ mm²/s as a threshold, we were able to differentiate malignant from benign lesions with 91% diagnostic accuracy, 90% sensitivity, 100% specificity, 100% positive predictive value and 57% negative predictive value.
Short-tau inversion-recovery echo-planar imaging (STIR-EPI) sequences applied to the chest-mediastinum provided potentially useful images for the differential diagnosis of benign and malignant lesions.
通过计算平均表观扩散系数（ADC）值，我们回顾性评估了弥散加权成像（DWI）与脂肪和背景信号抑制在鉴别胸部-纵隔良恶性病变中的作用。
34 例肺结节/纵隔肿块患者行胸部常规及 DWI 序列磁共振（MR）成像。所有患者均行 CT 检查。MR 成像后，患者行经胸 CT 引导下活检或纵隔镜检查。获得组织病理学诊断后，回顾性地将病变分为五组：腺癌（n=16）、鳞癌（n=12）、慢性肺炎（n=2）、恶性纵隔肿瘤（n=2）和典型类癌（n=2）。我们使用 Mann-Whitney U 检验比较了不同病变组的 ADC 值。
良性和恶性病变的 ADC 值之间存在统计学差异（p<0.05）。使用 1.25×10⁻³ mm²/s 作为 ADC 值的阈值，我们能够以 91%的诊断准确性、90%的敏感性、100%的特异性、100%的阳性预测值和 57%的阴性预测值区分恶性和良性病变。
短回波时间反转恢复回波平面成像（STIR-EPI）序列应用于胸部-纵隔，为鉴别良恶性病变提供了有潜在价值的图像。
