Group sparsity based classification for cervigram segmentation

Abstract

This paper presents an algorithm to classify pixels in uterine cervix images into two classes, namely normal and abnormal tissues, and simultaneously select relevant features, using group sparsity. Because of the large variations in image appearance due to changes of illumination, specular reflections and other visual noise, the two classes have a strong overlap in feature space, whether features are obtained from color or texture information. Using more features makes the classes more separable and increases the segmentation's quality, but also its complexity. However, the properties of these features have not been well investigated. In most cases, a group of features is selected prior to the segmentation process; features with minor contributions to the results are kept and add to the computational cost. We propose feature selection as a significant improvement in this problem. It provides a robust trade-off between segmentation quality and computational complexity. In this work we formulate the cervigram segmentation problem as a feature-selection-based classification method, and we introduce a regularization-based feature-selection algorithm to leverage both the sparsity and clustering properties of the features used. We implemented our method to automatically segment the biomarker AcetoWhite (AW) regions in a dataset of 200 images of the uterine cervix, for which manual segmentation is available. We compare the performance of several regularization-based feature-selection methods. The experimental results demonstrate that on this dataset, our proposed group-sparsity-based method gives overall better results in terms of sensitivity, specificity and sparsity.

Simultaneous Sparsity Model for Histopathological Image Representation and Classification

“…The robustness of sparse features to real-world imaging distortions has led to their widespread use in application domains such as remote sensing (hyperspectral imaging [28], synthetic aperture radar [29]) and biometrics (face recognition [27], [30]). SRC has also been proposed for single-channel medical images, in cervigram segmentation [31], [32] and colorectal polyp and lung nodule detection [33]. To the best of our knowledge, ours is the first discriminative sparsity model for multi-channel histopathological images.…”

“…Modifications to (3) include relaxing the nonconvex -term to the -norm [42] and introducing regularization terms to capture physically meaningful constraints [32].…”

“…The linear representation model alleviates this difficulty. That said, many recent approaches have employed SRC (or extensions) on image features (see [32] and [37] for example). Our classifier can be similarly employed on image features that are chosen based on domain knowledge.…”

Learning Based Segmentation of CT Brain Images: Application to Postoperative Hydrocephalic Scans

“…Recently, sparsity constrained learning methods have been developed for image classification [19] and found to be widely successful in medical imaging problems [20]–[23]. The essence of the aforementioned sparse representation based classification (SRC) is to write a test image (or patch) as a linear combination of training images collected in a matrix (dictionary), such that the coefficient vector is determined under a sparsity constraint.…”

SHIRC: A simultaneous sparsity model for histopathological image representation and classification

“…Additionally, the class-specific design of such dictionaries enables class assignment via a simple reconstruction errorbased metric. Sparsity-based classification schemes have also 2013 IEEE 10th International Symposium on Biomedical Imaging: From Nano to Macro San Francisco, CA, USA, April 7-11, 2013 978-1-4673-6455-3/13/$31.00 ©2013 IEEE been proposed for medical applications recently [10][11][12].…”