Color Doppler Ultrasound Improves Machine Learning Diagnosis of Breast Cancer
1
New York Medical College, Valhalla, NY 10595, USA
2
Department of Radiology, University of Pennsylvania, Philadelphia, PA 19104, USA
3
Department of Electrical and Systems Engineering, University of Pennsylvania, Philadelphia, PA 19104, USA
*
Author to whom correspondence should be addressed.
Diagnostics 2020, 10(9), 631; https://doi.org/10.3390/diagnostics10090631
Received: 6 July 2020
/
Revised: 20 August 2020
/
Accepted: 21 August 2020
/
Published: 25 August 2020
(This article belongs to the Section Machine Learning and Artificial Intelligence in Diagnostics)
Color Doppler is used in the clinic for visually assessing the vascularity of breast masses on ultrasound, to aid in determining the likelihood of malignancy. In this study, quantitative color Doppler radiomics features were algorithmically extracted from breast sonograms for machine learning, producing a diagnostic model for breast cancer with higher performance than models based on grayscale and clinical category from the Breast Imaging Reporting and Data System for ultrasound (BI-RADSUS). Ultrasound images of 159 solid masses were analyzed. Algorithms extracted nine grayscale features and two color Doppler features. These features, along with patient age and BI-RADSUS category, were used to train an AdaBoost ensemble classifier. Though training on computer-extracted grayscale features and color Doppler features each significantly increased performance over that of models trained on clinical features, as measured by the area under the receiver operating characteristic (ROC) curve, training on both color Doppler and grayscale further increased the ROC area, from 0.925 ± 0.022 to 0.958 ± 0.013. Pruning low-confidence cases at 20% improved this to 0.986 ± 0.007 with 100% sensitivity, whereas 64% of the cases had to be pruned to reach this performance without color Doppler. Fewer borderline diagnoses and higher ROC performance were both achieved for diagnostic models of breast cancer on ultrasound by machine learning on color Doppler features.
View Full-Text
Keywords:
ultrasound; color Doppler; radiomics; breast cancer; machine learning
►▼
Show Figures
---
Figure 2
Figure 3
This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited
-
Externally hosted supplementary file 1
Link: http://github.com/tedcary/UltraBoost
Description: This is a link to the UltraBoost Weka package which includes the zipped source code and binary. The package will work best with Weka 3.8.4 or above, though it will work with reduced functionality on earlier releases. To install the package manually to Weka on a local machine, download the ultraBoost.zip file from the link. From the Weka GUI Chooser menubar, select Tools->Package manager. Press the 'File/URL' button at the top right and navigate to the ultraBoost.zip file you downloaded, which will install the package. After installation, restart Weka. The UltraBoost classifier will now be available in the 'meta' classifiers. (In Weka Explorer, select the 'Classify' tab, press the 'Choose' button, and expand the tree to see weka->Classifiers->meta->UltraBoost.)
MDPI and ACS Style
Moustafa, A.F.; Cary, T.W.; Sultan, L.R.; Schultz, S.M.; Conant, E.F.; Venkatesh, S.S.; Sehgal, C.M. Color Doppler Ultrasound Improves Machine Learning Diagnosis of Breast Cancer. Diagnostics 2020, 10, 631. https://doi.org/10.3390/diagnostics10090631
AMA Style
Moustafa AF, Cary TW, Sultan LR, Schultz SM, Conant EF, Venkatesh SS, Sehgal CM. Color Doppler Ultrasound Improves Machine Learning Diagnosis of Breast Cancer. Diagnostics. 2020; 10(9):631. https://doi.org/10.3390/diagnostics10090631
Chicago/Turabian StyleMoustafa, Afaf F., Theodore W. Cary, Laith R. Sultan, Susan M. Schultz, Emily F. Conant, Santosh S. Venkatesh, and Chandra M. Sehgal. 2020. "Color Doppler Ultrasound Improves Machine Learning Diagnosis of Breast Cancer" Diagnostics 10, no. 9: 631. https://doi.org/10.3390/diagnostics10090631
Find Other Styles
Note that from the first issue of 2016, MDPI journals use article numbers instead of page numbers. See further details here.