探讨实时组织弹性成像技术在直肠肿瘤鉴别诊断中的应用价值。

选择2012年4月至2013年10月中国医科大学附属第一医院收治的经手术病理证实的直肠肿瘤患者100例，所有患者肿瘤均为单发。其中直肠腺瘤30例，直肠腺癌70例，并选择70名健康志愿者作为健康对照组。对所有受检者行经直肠超声检查。应用实时组织弹性成像技术对直肠肿瘤患者分别测量病灶与周围脂肪组织的应变率（SR）比值，对健康志愿者分别测量直肠肠壁与周围脂肪组织的SR比值；同时对其弹性成像图进行评分。绘制超声弹性评分法、SR比值法鉴别诊断直肠肿瘤的操作者工作特征（ROC）曲线，确定超声弹性评分法、SR比值法鉴别诊断直肠肿瘤的最佳界值，计算相应的敏感度、特异度、准确性。采用方差分析比较健康对照组受试者、直肠腺瘤患者、直肠腺癌患者SR比值差异、术前不同病理分期直肠腺癌患者SR比值差异，进一步组间两两比较采用LSD-t检验。

ROC曲线显示，超声弹性评分=3分为鉴别诊断直肠肿瘤的最佳界值，其鉴别诊断直肠肿瘤的敏感度为85.1%，特异度为73.1%，准确性为82.0%，曲线下面积为0.780；SR比值=2.34为鉴别诊断直肠肿瘤的最佳界值，其鉴别诊断直肠肿瘤的敏感度为91.4%，特异度为83.3%，准确性为89.0%，曲线下面积为0.945。健康对照组、直肠腺瘤患者、直肠腺癌患者SR比值分别为0.74±0.44、1.75±0.58、7.48±6.80；直肠腺癌患者SR比值均高于健康对照组、直肠腺瘤患者，且差异均有统计学意义（t = -8.26，P = 0.000；t= -6.98，P = 0.000）；直肠腺瘤患者SR比值高于健康对照组，且差异有统计学意义（t= -8.53，P = 0.000）。术前病理分期T₁、T₂、T₃、T₄期直肠腺癌患者SR比值分别为4.91±3.60、7.07±7.23、8.64±2.62、8.58±9.95，差异无统计学意义（F=0.86，P =0.47）。

实时组织弹性成像技术鉴别诊断直肠肿瘤具有较好的临床价值。

To evaluate the value of real-time ultrasonic elastography in the differential diagnosis of rectal tumors.

One hundred patients (30 cases of adenoma and 70 cases of adenocarcinoma) with rectal tumors proved by pathology and 70 normal subjects referred to the First Affiliated Hospital of China Medical University from April 2012 to October 2013 were included in this prospective cohort. All patients underwent real-time ultrasonic elastography. The tumour tissue and reference tissue were chosen for strain ratio measurements. Only tumor tissue was selected for the tumor sample area. Tissue recognized as normal rectal wall and perirectal tissue was selected as the reference sample area. At the same time the elasticity score were calculated. Receiver operating characteristic (ROC) curves of the elasticity score and strain ratio method were used to find the cut-off point of adenoma and adenocarcinoma. The corresponding sensitivity, accuracy and specificity were calculated. One-way ANOVA was used to compare strain ratio value among healthy control group, retal adenoma group and rectal adenocarcinoma group and among patients with different preoperative stages of adenocarcinoma group. LSD-t test was used to compare strain ratio value between two groups.

ROC curve showed that the best cut-off value of elasticity score in diagnosis of rectal tumor was 3, the sensitivity, specificity and accuracy were 85.1%, 73.1% and 82.0%, and the area under ROC curve was 0.780. The best cut-off value of strain ratio in diagnosis of rectal tumour was 2.34, the sensitivity, specificity and accuracy in diagnosis were 91.4%, 83.3% and 89.0%, and the area under ROC curve was 0.945. Strain ratio of healthy control group, rectal adenoma group and rectal adenocarcinoma group was 0.74±0.44, 1.75±0.58 and 7.48±6.80. There was significantly statistical difference among three groups in strain ratio. Compared with the strain ratio of healthy control group and rectal adenoma group, that of rectal adenocarcinoma group was higher (t=-8.26, P=0.000; t=-6.98, P=0.000). Compared with the strain ratio of healthy control group, that of rectal adenoma group was higher (t=-8.53, P=0.000). Strain ratio of patients with preoperative pathological stages T₁, T₂, T₃ and T4 rectal carcinoma was 4.91±3.60, 7.07±7.23, 8.64±2.62 and 8.58±9.95 and there was no significantly statistical difference among patients with different preoperative stages of adenocarcinoma group (F=0.86, P=0.47).

The real-time ultrasonic elastography is a promising modality for the differential diagnosis of rectal tumors.