Share Email Print
cover

Proceedings Paper

Breast image feature learning with adaptive deconvolutional networks
Author(s): Andrew R. Jamieson; Karen Drukker; Maryellen L. Giger
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

Feature extraction is a critical component of medical image analysis. Many computer-aided diagnosis approaches employ hand-designed, heuristic lesion extracted features. An alternative approach is to learn features directly from images. In this preliminary study, we explored the use of Adaptive Deconvolutional Networks (ADN) for learning high-level features in diagnostic breast mass lesion images with potential application to computer-aided diagnosis (CADx) and content-based image retrieval (CBIR). ADNs (Zeiler, et. al., 2011), are recently-proposed unsupervised, generative hierarchical models that decompose images via convolution sparse coding and max pooling. We trained the ADNs to learn multiple layers of representation for two breast image data sets on two different modalities (739 full field digital mammography (FFDM) and 2393 ultrasound images). Feature map calculations were accelerated by use of GPUs. Following Zeiler et. al., we applied the Spatial Pyramid Matching (SPM) kernel (Lazebnik, et. al., 2006) on the inferred feature maps and combined this with a linear support vector machine (SVM) classifier for the task of binary classification between cancer and non-cancer breast mass lesions. Non-linear, local structure preserving dimension reduction, Elastic Embedding (Carreira-Perpiñán, 2010), was then used to visualize the SPM kernel output in 2D and qualitatively inspect image relationships learned. Performance was found to be competitive with current CADx schemes that use human-designed features, e.g., achieving a 0.632+ bootstrap AUC (by case) of 0.83 [0.78, 0.89] for an ultrasound image set (1125 cases).

Paper Details

Date Published: 23 February 2012
PDF: 13 pages
Proc. SPIE 8315, Medical Imaging 2012: Computer-Aided Diagnosis, 831506 (23 February 2012); doi: 10.1117/12.910710
Show Author Affiliations
Andrew R. Jamieson, The Univ. of Chicago Medical Ctr. (United States)
Karen Drukker, The Univ. of Chicago Medical Ctr. (United States)
Maryellen L. Giger, The Univ. of Chicago Medical Ctr. (United States)


Published in SPIE Proceedings Vol. 8315:
Medical Imaging 2012: Computer-Aided Diagnosis
Bram van Ginneken; Carol L. Novak, Editor(s)

© SPIE. Terms of Use
Back to Top