Objective:To investigate the potential value of synthetic MRI is based on the ALN itself quantitative histogram features in predicting axillary lymph node metastasis(ALNM)of breast cancer,and to compare its predic-tive efficacy with conventional imaging features and ADC value quantitative histogram features.Methods:A total of 104 breast axillary lymph nodes(45 in the ALNM group and 59 in the non-ALNM group)were included.All patients underwent complete breast MR examination,including synthetic MRI,DWI and DCE MRI.PyRadiomics software was used to extract the quantitative histogram features of synthetic MRI and ADC.Chi-square test,independent sample t test and Mann-Whitney U test were used to compare the histogram features of synthetic MRI T1,T2 and proton den-sity(PD),ADC values and the differences of image signs on DCE-MRI between the ALNM group and the non-ALNM group.The parameters with statistically significant differences were included in multivariate logistic regression analysis,and the area under the curve(AUC)of ROC was used to evaluate the diagnostic performance of the above variables in predicting axillary lymph node metastasis status of breast cancer.Results:The T2-minimum value,ADC-10th and lymph node DCE-MRI image sign(short diameter)in synthetic MRI were independent predictors for differentiating axillary lymph node metastasis of breast cancer(P values were 0.041,0.022,0.023,respec-tively).However,the prediction model established by T2-minimum value has better diagnostic efficiency[0.765(95％CI:0.673 to 0.858)]and was superior to ADC(ADC-10th)and image sign(short diameter)prediction models.In addition,the multi-parameter model established by synthetic MRI(T2-minimum)+ADC(ADC-10th)+image sign(short diameter)significantly improved the overall diagnostic efficiency of axillary lymph node metastasis of breast cancer.Conclusion:The Synthetic MRI is based on the ALN itself quantitative parameter histo-gram features have certain value in predicting the status of axillary lymph node metastasis of breast cancer,and the synthetic MRI+ADC quantitative parameter histogram features+lymph node image signs multi-parameter model has the best prediction efficiency.This multi-parameter combined model may be a promising alternative method for predicting ALNM.
breast cancersynthetic MRIlymph nodesADChistogram features
维普
万方数据
