论文部分内容阅读
目的探索并建立乳腺癌血瘀证量化诊断标准,实现诊断过程可视化、判别过程自动化。方法收集乳腺癌回顾性病历资料60例,用以建立乳腺癌血瘀证量化诊断标准及对制定标准的回顾性检验。收集乳腺癌住院患者病历资料31例,对制定的乳腺癌血瘀证量化诊断标准进行前瞻性检验。Logistic回归获得乳腺癌血瘀证相关证候条目,条件概率换算法对相关证候条目赋值,ROC曲线法获得最佳诊断阈值,JAVA语言编程实现判别过程的自动化。结果获得乳腺癌血瘀证量化诊断标准,诊断阈值为16分。量化诊断标准前瞻性检验敏感度95.0%、特异度81.8%、准确度90.3%、阳性似然比5.22%、阴性似然比6.11%。获得乳腺癌血瘀证量化诊断软件。结论制定的乳腺癌血瘀证量化诊断标准适合临床运用,可结合软件进行推广。
Objective To explore and establish a quantitative diagnostic criteria of blood stasis syndrome in breast cancer to visualize the diagnosis process and to make the discrimination process automation. Methods Sixty cases of retrospective medical records of breast cancer were collected to establish the quantitative diagnostic criteria of blood stasis syndrome of breast cancer and the retrospective test of the standard. 31 cases of breast cancer inpatient records were collected and prospectively tested for the quantitative diagnostic criteria of blood stasis syndrome of breast cancer. Logistic regression was used to obtain syndromes related to blood stasis syndrome in breast cancer. Conditional probability method was used to evaluate the related syndromes, ROC curve method was used to get the best diagnosis threshold, and JAVA language was used to automate the discrimination process. Results obtained quantitative diagnosis of breast cancer blood stasis diagnostic criteria, the diagnostic threshold of 16 points. Quantitative diagnostic criteria for the prospective test sensitivity of 95.0%, specificity of 81.8%, accuracy of 90.3%, positive likelihood ratio of 5.22%, negative likelihood ratio of 6.11%. Get blood stasis syndrome quantitative diagnostic software. Conclusions The quantitative diagnostic criteria for developing blood stasis syndrome of breast cancer is suitable for clinical application and can be popularized in combination with software.