Abstract
The goal of this research was to optimize performance of a Computer-Aided Diagnostic system to identify, analyze and compare breast masses based on parameters measured in the ultrasound image. We compared case-based reasoning using Relative Similarity to an Artificial Neural Network in order to implement an objective form of the ACR BIRADS scheme to describe and score breast masses. The image feature set was reduced to nine including margins, shape, echogenicity, echo texture, orientation and posterior acoustic attenuation. Both classifiers performed well with a high ROC AZ although RS performed significantly better than the ANN in Specificity, PPV and achieved the goal of very high Specificity without a reduction in Sensitivity. Compared to a preliminary version of the RS classifier this optimized version of RS has significantly higher AZ (0.96 vs. 0.93)
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Stavros AT, Thickman D, Rapp CL, Dennis MA, et al., “Solid Breast Nodules: Use of Sonography to Distinguish between Benign and Malignant Lesions,” Radiology 196, pp. 123–134, 1995.
American College of Radiology: ACR Standards 2000–2001. Reston, VA: American College of Radiology, 2000.
André MP, Galperin M, Olson LK, et al, “Preliminary investigation of a method to assess breast ultrasound level of suspicion,” SPIE Medical Imaging 4322:507–512, 2001.
André MP, Galperin M, Olson LK, et al.: “Improving the accuracy of diagnostic breast ultrasound,” Acoustical Imaging 26:453–460, 2002.
Andre MP, Galperin M, Phan P, Chiu P, “ROC analysis of lesion descriptors in breast ultrasound images,” SPIE Medical Imaging 5034, 2003.
André MP, Galperin M, Green CT, Olson L: “A Case-Based Reasoning System to Aid Interpretation of Breast Ultrasound Images”, SPIE Medical Imaging 5749, 2003.
Mackay, DJC. (1992): A practical Bayesian framework for back-propagation networks. Neural Computation, 4(3), 448–472.
Mhitaryan V, Mayzlin I, Troshin L: “Determining the optimal decisions points for the parameters of the classified objects,” Methods of Computational Mathematics and its Application, Institute of Economics and Statistics, Moscow, 1985.
Kupinski MA, Giger ML, “Feature selection with limited dataset,” Med. Phys. 26(10), pp. 2176–2182, 1999.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2007 Springer
About this paper
Cite this paper
Andrè, M., Galperin, M., Contro, G., Omid, N., Olson, L. (2007). Optimization of a Breast Mass Classifier for Computer-Aided Ultrasound Analysis. In: André, M.P., et al. Acoustical Imaging. Acoustical Imaging, vol 28. Springer, Dordrecht. https://doi.org/10.1007/1-4020-5721-0_28
Download citation
DOI: https://doi.org/10.1007/1-4020-5721-0_28
Publisher Name: Springer, Dordrecht
Print ISBN: 978-1-4020-5720-5
Online ISBN: 978-1-4020-5721-2
eBook Packages: Physics and AstronomyPhysics and Astronomy (R0)