Please use this identifier to cite or link to this item:

http://hdl.handle.net/10609/93200
Title: Automated breast ultrasound lesions detection using convolutional neural networks
Author: Yap, Moi Hoon
Pons Rodríguez, Gerard
Martí Bonmatí, Joan
Ganau Macías, Sergi
Sentís Crivellé, Melcior
Zwiggelaar, Reyer
Davison, Adrian K.
Martí Marly, Robert
Others: Manchester Metropolitan University
Universitat de Girona
Aberystwyth University
University of Manchester
Universitat Oberta de Catalunya (UOC)
Keywords: breast cancer
convolutional neural networks
lesion detection
transfer learning
ultrasound imaging
Issue Date: Jul-2018
Publisher: IEEE Journal of Biomedical and Health Informatics
Citation: Yap, M.H., Pons, G., Martí, J., Ganau, S., Sentís, M., Zwiggelaar, R., Davison, A.K. & Martí, R. (2018). Automated breast ultrasound lesions detection using convolutional neural networks. IEEE Journal of Biomedical and Health Informatics, 22(4), 1218-1226. doi: 10.1109/JBHI.2017.2731873
Also see: https://doi.org/10.1109/jbhi.2017.2731873
Abstract: Breast lesion detection using ultrasound imaging is considered an important step of computer-aided diagnosis systems. Over the past decade, researchers have demonstrated the possibilities to automate the initial lesion detection. However, the lack of a common dataset impedes research when comparing the performance of such algorithms. This paper proposes the use of deep learning approaches for breast ultrasound lesion detection and investigates three different methods: a Patch-based LeNet, a U-Net, and a transfer learning approach with a pretrained FCN-AlexNet. Their performance is compared against four state-of-the-art lesion detection algorithms (i.e., Radial Gradient Index, Multifractal Filtering, Rule-based Region Ranking, and Deformable Part Models). In addition, this paper compares and contrasts two conventional ultrasound image datasets acquired from two different ultrasound systems. Dataset A comprises 306 (60 malignant and 246 benign) images and Dataset B comprises 163 (53 malignant and 110 benign) images. To overcome the lack of public datasets in this domain, Dataset B will be made available for research purposes. The results demonstrate an overall improvement by the deep learning approaches when assessed on both datasets in terms of True Positive Fraction, False Positives per image, and F-measure.
Language: English
URI: http://hdl.handle.net/10609/93200
ISSN: 2168-2194MIAR

2168-2208MIAR
Appears in Collections:Articles
Articles

Share:
Export:
Files in This Item:
File SizeFormat 
automatedultrasound.pdf896.61 kBAdobe PDFView/Open Request a copy

Items in repository are protected by copyright, with all rights reserved, unless otherwise indicated.