Differentiation of benign and metastatic axillary lymph nodes in breast cancer: additive value of MRI computer-aided evaluation. Issue 4 (April 2016)

• Record Type: Journal Article
• Authors: Yun, S.J.; Sohn, Y.-M.; Seo, M.
• Is Part Of: Clinical radiology. Volume 71:Issue 4(2016)
• Page Start: 403.e1
• Page End: 403.e7
• Publication Date: 2016-04
• Subjects: Medical radiology -- Periodicals; Radiotherapy -- Periodicals; Radiotherapy -- Periodicals; Radiology -- Periodicals; Societies, Medical -- Periodicals; Medical radiology; Radiotherapy; Electronic journals; Periodicals; 616.0757
• DOI: 10.1016/j.crad.2016.01.008
• Languages: English
• ISSNs: 0009-9260

Abstract

Abstract : Aim: To retrospectively determine the qualitative and quantitative cut-off values of the magnetic resonance computer-aided evaluation (MR CAE) parameters to differentiate between benign and metastatic axillary lymph nodes (ALNs) and to investigate the combined diagnostic performance of MR CAE imaging. Materials and methods: From July 2011 to June 2014, 124 patients who underwent preoperative conventional MR, diffusion-weighted (DW), and MR CAE imaging were included. Computer-generated qualitative and quantitative features for ALNs were recorded, and two breast imaging radiologists interpreted the MR images for the presence of metastatic ALNs using conventional MR and DW, and in combination with MR CAE images by consensus. The cut-off values of MR CAE and diagnostic performance were derived from the receiver operating characteristic (ROC) curve. Results: Thirty-four (26.4%) were ALN positive and 90 (73.6%) were ALN negative on the final histopathological result. On qualitative analysis, visualization on the colour map ( p =0.007) and kinetic curve type ( p <0.001) were significantly different between the groups. On quantitative analysis, mean values (%) of persistent, plateau, and washout ratios differed significantly ( p <0.001). Of these significant parameters, a washout ratio of >49% showed the greatest diagnostic accuracy (area under the ROC curve, 0.909). With conventional MR and DW images, sensitivity, specificity, and accuracy were 82.4%, 85.6%, and 84.7%, respectively. With added information from MR CAE images, accuracy significantly improved to 93.5% ( p =0.043). The sensitivity and specificity improved to 91.2% ( p =0.403) and 94.4% ( p =0.086), respectively. Conclusion: The additive use of MR CAE improved diagnostic performance and can be helpful for differentiating benign from metastatic ALNs. Highlights: Among MR CAE parameters, washout ratio>49% showed the greatest diagnostic accuracy. With added information from MR CAE images, accuracy significantly improved to 93.5%. The MR CAE can be helpful for differentiating benign from metastatic ALNs.