Unsupervised feature selection applied to content-based retrieval of lung images

Abstract

Abstract-This paper describes a new hierarchical approach to content-based image retrieval called the "customized-queries" approach (CQA). Contrary to the single feature vector approach which tries to classify the query and retrieve similar images in one step, CQA uses multiple feature sets and a two-step approach to retrieval. The first step classifies the query according to the class labels of the images using the features that best discriminate the classes. The second step then retrieves the most similar images within the predicted class using the features customized to distinguish "subclasses" within that class. Needing to find the customized feature subset for each class led us to investigate feature selection for unsupervised learning. As a result, we developed a new algorithm called FSSEM (feature subset selection using expectation-maximization clustering). We applied our approach to a database of high resolution computed tomography lung images and show that CQA radically improves the retrieval precision over the single feature vector approach. To determine whether our CBIR system is helpful to physicians, we conducted an evaluation trial with eight radiologists. The results show that our system using CQA retrieval doubled the doctors' diagnostic accuracy.

Spectral clustering with eigenvector selection

“…However, in practice, the inclusion of uninformative eigenvectors can degrade the clustering process as demonstrated extensively later in the paper. This is hardly surprising because in a general context of pattern analysis, the importance of removing those noisy/uninformative features has long been recognised [2,5]. The answer to the second question is thus 'yes'.…”

Unsupervised Gene Selection and Clustering Using Simulated Annealing

“…Most of the literature on feature selection pertains to supervised learning, and not much work has been done for feature selection in unsupervised learning [13,6,11,8,14,2].…”

Feature Selection (Unsupervised Learning)

“…We examine the FSSEM variants on the iris, wine, and ionosphere data set from the UCI learning repository (Blake & Merz, 1998), and on a high resolution computed tomography (HRCT) lung image data which we collected from IUPUI medical center (Dy et al, 2003;Dy et al, 1999). Although for each data set the class information is known, we remove the class labels during training.…”