摘要
目的:分析卵巢囊实性病变的MRI表现,提高诊断水平。方法:搜集44例经手术病理证实的卵巢囊实性病变的临床及MRI资料,分析影像表现,采用统计学方法,建立判别函数,对其MRI征象诊断价值做进一步判断。结果:卵巢良、恶性病变在囊壁是否规则、有无壁结节、与邻近组织分界是否清晰、盆腔有无肿大淋巴结及腹水等表现差异有统计学意义,P<0.05。经Bayes逐步判别,经F检验,囊壁是否规则、有无壁结节、与邻近组织分界是否清晰等变量对肿瘤的良、恶性有显著判别能力,以此建立判别函数,总体诊断准确度为93.18%。结论:MRI对卵巢病变的诊断有重要价值,判别函数能进一步提高诊断的准确性。
Objective：To analvse the MRI manifestations of ovarian cystic and solid mass.Methods：Forty-four cases with pathological confirmed ovarian cystic lesions were enrolled in this study.For all of the patients,routine MRI was performed.Image findings of the masses were analyzed using discriminant function.Results：Significant differences between benign tumors and malignant tumors were found in regular or irregular wall,with or without mural nodule,sharp or unclear edge,with or without swelling pelvic nodule and with or without ascites （P＜0.05）.Discriminant function analysis using these regular or irregular wall,with or without mural nodule,and markers of edge was used to identify discriminatory variables using a stepwise procedure.The function classified overall 93.18％ of the cases correctly.Conclusion：MRI may be used to differentiate between benign ovarian tumors and malignant tumors.Using discriminant functions could further improve the accuracy of the diagnosis.
出处
《中国临床医学影像杂志》
CAS
2014年第2期106-108,共3页
Journal of China Clinic Medical Imaging
关键词
卵巢肿瘤
磁共振成像
Ovarian neoplasms
Magnetic rsonance imaging
