Objective To systematically evaluate the clinical value of ultrasound radiomics in the diagnosis of axillary lymph node metastasis(ALNM)in patients with breast cancer by Meta-analysis.Methods PubMed,Web of Science,Embase,Cochrane Library,CNKI,Wanfang Data knowledge service platform,VIP database and China biomedical literature database were searched by computer.The databases were conducted to collect diagnostic studies on ultrasound radiomics for ALNM in breast cancer patients.The search time limit was set from inception to October 1,2023.Two researchers screened the literature and extracted valid data,statistical software Stata 17.0,RevMan 5.3 and Meta-Disc 1.4 were used for Meta-analysis.Results A total of 14 literatures were included in this study,there were 4469 cases of breast cancer patients,including 1748 cases of ALNM patients and 2721 cases of non-ALNM patients.The results of Meta-analysis showed that the combined sensitivity of ultrasound radiomics for ALNM diagnosis in breast cancer patients was 0.79[95%confidence interval(CI):0.73～0.85],the combined specificity was 0.79(95%CI:0.71～0.85),the combined positive likelihood ratio was 3.70(95%CI:2.70～5.10),the combined negative likelihood ratio was 0.26(95%CI:0.19～0.35),the combined diagnostic odds ratio was 14.0(95%CI:8.0～24.0),and the area under the comprehensive receiver operating characteristic curve was 0.86(95%CI:0.83～0.89).The heterogeneity of the combined results was not caused by a single study.Subgroup analysis showed that the sample size and the publication area of the literature might be the reasons for the heterogeneity.Deek's funnel plot showed that there was no potential publication bias(t=0.37,P=0.72).Fagan plot showed that the prior probability of ultrasound radiomics in the diagnosis of ALNM was 50%,when the positive likelihood ratio was 4 and the negative likelihood ratio was 0.26,the posterior probability was 79%and 21%,respectively.Conclusion Ultrasound radiomics has a certain clinical value in the diagnosis of ALNM in patients with breast cancer.
NETL
NSTL
万方数据
