Radiomic features of axillary lymph nodes based on pharmacokinetic modeling DCE-MRI allow preoperative diagnosis of their metastatic status in breast cancer

• DOI: 10.1371/journal.pone.0247074
• Publication Date: 2021-03-01T21:03:52Z
• Authors: Hong-Bing Luo; Yuan-Yuan Liu; Chun-hua Wang; Hao-Miao Qing; Min Wang; Xin Zhang; Xiao-Yu Chen; Guo-Hui Xu; Peng Zhou; Jing Ren
• Keywords: Axillary lymph nodes; Axilla

To study the feasibility of use radiomic features extracted from axillary lymph nodes for diagnosis their metastatic status in patients with breast cancer.A total 176 cancer, consisting 87 (ALNM) and 89 negative proven by surgery, were retrospectively reviewed database our cancer center. For each selected node, 106 based on preoperative pharmacokinetic modeling dynamic contrast enhanced magnetic resonance imaging (PK-DCE-MRI) 5 conventional image obtained. The least absolute shrinkage selection operator (LASSO) regression was used to select useful features. Logistic develop diagnostic models ALNM. Delong test compare performance different models.The reduced 4 ALNM diagnosis-related LASSO. Four including model, a combined model (integrating Rad-score features) developed validated. showed that had best performance: area under curve (AUC), 0.972 (95% CI [0.947-0.997]) training cohort 0.979 [0.952-1]) validation cohort. better than (P<0.05).Radiomic PK-DCE-MRI images promising application cancer.