Study On Preoperative Prediction Of Lymph Node Metastasis In Uyghur Triple-Negative Breast Cancer Based On MRI Radiomics Characteristics

Objective: To investigate the value of preoperative imaging radiomics characteristics for Uyghur triple-negative breast cancer(TNBC)axillary lymph node(ALN)metastasis based on dynamic contrast-enhanced magnetic resonance imaging(DCE-MRI)and diffusion-weighted imaging(DWI).Methods: A retrospective analysis of the clinicopathology and MRI data of 120 Uyghur TNBC patients confirmed from January2013 to December 2021.They were divided into 86 cases(32 ALN positive,54 negative)in training set and 34 cases(16 ALN positive and 18 negative)according to the ratio of7:3,image segmentation and imaging radiomics feature extraction for manually delineated ROI from DCE-MRI Phase I and DWI(b=800),Multivariate Logistics regression analysis was used to establish the imaging radiomics prediction model based on DCE-MRI,DWI and DCE-MRI and DWI.The diagnostic efficacy was evaluated by the receiver operating characteristic curve(ROC);Results: The imaging radiomics prediction model of DCE-MRI,DWI and DCE-MRI combined with DWI were 0.83,0.80 and 0.87 in the training group;the AUC of the validation group was 0.73,0.74,and 0.79,respectively.The DCE-MRI combined DWI model showed the best predictive efficacy in the three models.Conclusion: The imaging radiomics prediction model based on DCE-MRI and DWI has a certain value to predict the Uyghur TNBC axillary lymph node metastasis status preoperative.The diagnostic efficiency of the combined model of DCE-MRI and DWI is better than the single sequence model,and the 10 selected imaging radiomics features can predict Uyghur TNBC lymph node status noninvasively and accurately.