Non-Mass Breast Lesions on Ultrasound: Feature Exploration and Multimode Ultrasonic Diagnosis.

Ultrasound Med Biol

Department of Ultrasound, Sun Yat-sen Memorial Hospital, Sun Yat-sen University, Guangzhou, China; Guangdong Province Key Laboratory of Malignant Tumor Epigenetics and Gene Regulation, Sun Yat-Sen Memorial Hospital, Sun Yat-Sen University, Guangzhou, China. Electronic address:

Published: August 2018

The aim of this study was to analyze the features of non-mass breast lesions (NMLs) on B-mode ultrasound (US), color Doppler US, strain elastography (SE) and contrast-enhanced ultrasound (CEUS) and to develop a multimode ultrasonic method for NML differentiation. Seventy-one NMLs were included in this retrospective study. Binary logistic regression was used to identify the independent risk factors. Pathology results were used as the standard criterion. Microcalcification on US, high stiffness on SE and hyper-enhanced intensity on CEUS were identified as features correlated with malignancy. A multimode method to evaluate NMLs based on the logistic regression was developed. The sensitivity and specificity for US, US + Doppler, US + SE, US + CEUS and the multimode method were 100% and 29%, 92.5% and 41.9%, 97.5% and 58.1%, 90.0% and 58.1% and 95.0% and 77.4%, respectively. The accuracy of these methods was 69.0%, 70.4%, 80.2%, 76.1% and 87.3%, respectively. The multimode ultrasonic method is simple and exhibited high diagnostic performance, which might be helpful for predicting the potential malignancy of NMLs.

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.ultrasmedbio.2018.05.005DOI Listing

Publication Analysis

Top Keywords

multimode ultrasonic
12
non-mass breast
8
breast lesions
8
ultrasonic method
8
logistic regression
8
multimode method
8
multimode
5
lesions ultrasound
4
ultrasound feature
4
feature exploration
4

Similar Publications

Want AI Summaries of new PubMed Abstracts delivered to your In-box?

Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!