卵巢恶性肿瘤的危险因素在彩色超声中的表现

发布时间：2018-02-10 12:20

本文关键词： 卵巢恶性肿瘤超声多普勒危险因素　出处：《吉林大学》2014年硕士论文　论文类型：学位论文

【摘要】：目的：探讨临床193例卵巢肿瘤患者的彩色超声图像特征中能够鉴别诊断出恶性肿瘤的危险因素有哪些，由此建立一个能单独运用彩色超声鉴别诊断卵巢恶性肿瘤的危险分级，并检验此危险分级的临床应用价值。方法：所有患者选择于2012年1月至2012年12月期间经吉林大学白求恩第一医院妇科诊治的193例卵巢肿瘤患者，术前均由我院妇科超声检查室行彩色多普勒超声检查，术后均由我院病理科行常规病理检查，并依照FIGO的手术-病理分期进行最终诊断。不考虑每例卵巢肿瘤的病理诊断结果，仅对其术前彩色多普勒超声图像进行回顾性分析，共总结出所有卵巢肿瘤的11种彩色超声特征表现。利用统计学二元logistic回归，将卵巢恶性和交界性肿瘤作为一组，将卵巢良性肿瘤作为一组，识别出与卵巢恶性及交界性肿瘤最相关的彩色多普勒超声图像特征，由此建立一能够提示恶性及交界性肿瘤风险的三级危险分层。结果：根据二元logistic回归，得出与卵巢恶性或交界性肿瘤相关的高危、中危和低危因素：高危因素（即提示卵巢肿瘤为恶性或交界性的风险最高）是卵巢肿瘤含有内部血流；中危因素（即提示卵巢肿瘤为恶性或交界性的风险次之高危因素）是卵巢肿瘤为厚壁、厚分隔及含有实性或不规则区域；低危因素（即提示卵巢肿瘤为恶性或交界性的风险较低）是卵巢肿瘤为单纯囊肿、薄分隔、较稠回声及低回声等。将此危险分层于对照组中进行检验，假设包含高危或中危因素者为恶性或交界性，假设仅包含低危因素者认为是良性，与最终与病理结果对比，其鉴别出恶性及交界性的敏感度、特异度、阳性及阴性预测值分别为92.4%、58.2%、53.5%及93.7%。假设以仅包含高危因素者认为恶性或交界性，假设包含中危或低危因素者为良性，则得到的敏感度、特异度、阳性及阴性预测值分别为63.6%、81.1%、65%及81.1%。结论：二元logistic回归所建立的恶性肿瘤风险的三级危险分层可应用于临床以评估卵巢肿瘤为恶性或交界性的风险。对于卵巢肿瘤患者，，若其彩色超声结果能排除高危及中危因素，那么其为恶性或交界性肿瘤的几率较低，临床可行保守的随访观察。
[Abstract]:Objective: to study the risk factors of malignant tumor in 193 patients with ovarian tumor by color ultrasound, and to establish a risk grade which can be used to differentiate ovarian malignant tumor by color ultrasound alone. The clinical application value of the risk classification was tested. Methods: from January 2012 to December 2012, 193 patients with ovarian neoplasms were examined by color Doppler ultrasound in gynecological examination room of our hospital from January 2012 to December 2012 in the first Hospital of Bai Qiuen, Jilin University. Routine pathological examination was performed by pathology department of our hospital after operation, and the final diagnosis was made according to FIGO's operation-pathological stage. Without considering the pathological diagnosis results of each case of ovarian tumor, the preoperative color Doppler ultrasound images were analyzed retrospectively. 11 color ultrasound features of all ovarian tumors were summarized. Using statistical binary logistic regression, malignant and borderline ovarian tumors were selected as a group and benign ovarian tumors as a group. The color Doppler imaging features most associated with ovarian malignant and borderline tumors were identified, and a third level of risk stratification was established to indicate the risk of malignant and borderline tumors. Results: according to the binary logistic regression, the high risk factors associated with ovarian malignant or borderline tumors, middle and low risk factors: the high risk factors (that is, the highest risk of malignant or borderline ovarian tumors) were found to be the ovarian tumors containing internal blood flow. Middle risk factors (indicating that ovarian tumors are malignant or borderline) are thick-walled, thick-spaced and contain solid or irregular areas. The low risk factor (which indicates that the ovarian tumor is malignant or borderline) is that the ovarian tumor is simple cyst, thin septum, thicker echo and low echo. Those with high or moderate risk factors were assumed to be malignant or borderline, and those with only low risk factors were considered benign. Compared with the final pathological results, the sensitivity and specificity of the differentiation between malignant and borderline factors were found. The positive and negative predictive values were 92.4% and 93.7%, respectively. The sensitivity, specificity, positive and negative predictive values were 63.6%, 81.65% and 81.1%, respectively, assuming that those with high risk factors were considered malignant or borderline, and those with moderate or low risk factors were benign. Conclusion: the risk stratification of malignant tumors established by binary logistic regression can be used to evaluate the risk of ovarian tumors being malignant or borderline. If the results of color ultrasound can exclude the high and middle risk factors, the probability of malignant or borderline tumor is lower, and the clinical observation is feasible and conservative.
【学位授予单位】：吉林大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：R737.31;R445.1

【共引文献】