判别函数在腮腺肿瘤CT诊断中的价值

目的：分析腮腺肿瘤的CT表现特征,探讨判别函数在腮腺良恶性肿瘤诊断中的价值。方法：回顾性分析85例经病理证实的腮腺肿瘤病例资料。对良、恶性肿瘤组间的性别、部位、边界、平扫密度及颈部淋巴结肿大等指标行χ^2检验;对年龄、横断面最大径及CT增强幅值等指标行两独立样本t检验。根据肿瘤最大径、边界、CT增强幅值及颈部淋巴结肿大等指标建立判别函数。结果：良恶性肿瘤间患者年龄、平扫边界、横断面最大径、增强幅值及颈部淋巴结等指标的差异有统计学意义;而性别和肿瘤发生部位的差异无统计学意义。通过建立判别函数,对良、恶性肿瘤的判别符合率分别为93.8%和90.0%,整体的判别符合率为92.9%。结论：以肿瘤边界、横断面最大径、增强幅值及颈部淋巴结等四项指标建立判别函数对良、恶性腮腺肿瘤的鉴别能力强。

Value of Discriminant Function Applied in the CT Diagnosis of Parotid Tumors

Objective： To Analyse the CT manifestations of parotid tumors and to study the value of discriminant function applied in the CT diagnosis of parotid tumors. Methods： The clinical and CT materials of 85 patients with parotid tumors were collected. Pearson Chi Square （χ^2） test was used to compare those parameters including gender, tumor location,border,attenuation on plain scan and lymph adenopathy of neck between benign and malignant tumors. Age,maximum diameter in axial plane and CT value after contrast enhancement were compared by Independent Samples T test. Discriminant function was established according to various parameters such as tumor maximum diameter, border, attenuation value after enhancement and lymph adenopathy of neck. Results： Statistical significant differences were found in age, maxmum tumor diameter in axial plane, border, CT value after enhancement and lymph-adenopathy of neck between benign and malignant tumors. But no statistical significant differences were found in gender, tumor location and attenuation on plain scan. The accuracy of differential diagnosis of benign and malignant parotid tumors was 93.8% and 90.0% respectively by the application of value of discriminant function. Totally the accuracy of discriminant was 92.9%. Conclusion.. When tumor border, maximum diameter on axial plane,CT value after enhancement and adenopathy of neck were used to establish the discriminant function,it showed significant capability in differentiating benign and malignant parotid tumors.