CESM量化特征联合形态学对BI-RADS 3~5类乳腺单发肿块的诊断价值

Diagnostic value of quantitative features of CESM combined with morphology for single breast masses in BI-RADS category 3 to 5

目的:探讨对比增强能谱乳腺摄影(contrast enhancement spectral mammography,CESM)量化特征联合形态学对乳腺单发肿块的诊断价值。方法:回顾并分析2018年1月—2020年12月行CESM检查的乳腺影像报告和数据系统(Breast Imaging Reporting and Data System,BI-RADS)3~5类乳腺单发肿块。所有患者均行头尾(craniocaudal,CC)位(早期,_(T1))、内外斜(mediolateral oblique,MLO)位(晚期,_(T2))摄影,随后测量减影图病灶的强化灰度值(lesion grey value,LGV)。并根据LGV在两个时相的变化量评估病灶的动态强化特点。同时分析病灶CESM低能图及减影图形态学特征。最后以病理学检查结果为金标准,评价CESM(低能图及减影图)、LGV_(T1)+LGV_(T2)+动态强化、CESM+LGV_(T1)+LGV_(T2)、CESM+动态强化、CESM+LGV_(T1)+LGV_(T2)+动态强化对乳腺单发肿块的诊断效能。结果:160例患者全为女性,共检出160个肿块,恶性病灶118个,良性肿块42个。CESM、LGV_(T1)+LGV_(T2)+动态强化、CESM+LGV_(T1)+LGV_(T2)、CESM+动态强化及CESM+LGV_(T1)+LGV_(T2)+动态强化的曲线下面积(area under curve,AUC)分别为0.917、0.748、0.922、0.919及0.927。单独CESM诊断的灵敏度、特异度、阳性预测值、阴性预测值及准确度均高于LGV_(T1)+LGV_(T2)+动态强化(分别为91.5%vs 75.4%、83.3%vs 64.2%、93.9%vs 85.5%、77.7%vs 48.2%和89.3%vs 72.5%)。而CESM+LGV_(T1)+LGV_(T2)+动态强化的灵敏度、特异度、阳性预测值、阴性预测值及准确度分别为96.6%、80.9%、93.4%、89.4%、92.5%。结论:CESM(低能图+减影图)形态学、病灶的LGV及动态强化特征综合判断可提高BI-RADS 3~5类乳腺单发肿块的诊断效能,在一定程度上降低良性肿块的活检率。

Objective:To find out the benefits of adding morphology and quantitative analysis onto contrast enhancement spectral mammography(CESM)in diagnosis of single breast masses.Methods:All patients found with Breast Imaging Reporting and Data System(BI-RADS)category 3-5 breast mass-like lesions by CESM from January 2018 to December 2020 were retrospective analyzed.All cases were performed craniocaudal(CC)view(early phase,_(T1))and mediolateral oblique(MLO)view(later phase,_(T2)).The quantitative enhanced lesion gray value(LGV)was measured within each phase’s recombined image.Based on their changes,the kinetic enhancement pattern was assessed among the two phases.And the diagnostic performance was subsequently measured.Results:A total of 160 breast masses were found in 160 female patients,and 118 masses were malignant and 42 masses were benign.The areas under curves(AUC)of CESM,LGV_(T1)+LGV_(T2)+kinetic enhancement,CESM+LGV_(T1)+LGV_(T2),CESM+kinetic enhancement,and CESM+LGV_(T1)+LGV_(T2)+kinetic enhancement was 0.917,0.748,0.922,0.919 and 0.927,respectively.By evaluating the diagnostic value,sensitivity,specificity,positive predictive value,negative predictive value,and accuracy from morphological categorization were all higher than those of the LGV combined with kinetic enhancement(91.5%vs 75.4%,83.3%vs 64.2%,93.9%vs 85.5%,77.7%vs 48.2%,and 89.3%vs 72.5%).While,the sensitivity,specificity,positive predictive value,negative predictive value,and accuracy of the CESM+LGV_(T1)+LGV_(T2)+kinetic enhancement were 96.6%,80.9%,93.4%,89.4%and 92.5%,respectively.Conclusion:For the BI-RADS 3-5 breast masses,adding the CESM quantitative analysis with the CESM morphology can improve confidence level,and reduce the biopsy rate of benign masses.