Quantitative ultrasound image analysis of axillary lymph node status in breast cancer patients.

摘要: Ultrasonography has the potential to accurately stage breast cancer with automated analysis detect axillary lymph node metastasis. The aim of this study was develop and test quantitative ultrasound image nodes for staging. Following an IRB-approved HIPAA compliant protocol, images 90 patients presenting assessment were retrospectively collected. There 51 node-positive 39 node-negative patients, yielding 223 (109 positive metastasis 114 negative metastasis). completely apart from manual indication approximate center each node. Mathematical descriptors nodes, which served as image-based biomarkers, computer-extracted input a classifier task distinguishing between (i.e., metastatic) nodes. performance assessed using receiver operating characteristic (ROC) evaluation by-node by-patient area under ROC curve (AUC) metric. AUC 0.85 (standard error 0.03) when 0.87 (0.04) patient-based prognosis, i.e., assessing whether or node-negative. Based on these classification results, we conclude that mathematical sonographically imaged may be useful prognostic biomarkers in staging demonstrate predicting patient status.