Computer-assisted pit-pattern classification in different wavelet domains for supporting dignity assessment of colonic polyps

Abstract

In this paper, we show that zoom-endoscopy images can be well classified according to the pit-pattern classification scheme by using texture-analysis methods in different wavelet domains. We base our approach on three different variants of the wavelet transform and propose that the color-channels of the RGB and LAB color model are an important source for computing image features with high discriminative power. Color-channel information is incorporated by either using simple feature vector concatenation and cross-cooccurrence matrices in the wavelet domain. Our experimental results based on k-Nearest Neighbor classification and forward feature selection exemplify the advantages of the different wavelet transforms and show that color-image analysis is superior to grayscale image analysis regarding our medical image classification problem.

Automated Marsh-like classification of celiac disease in children using local texture operators

“…• DT-CWT Correlation Signatures: We have extended the Wavelet Correlation Signatures approach of de Wouwer et al (1997) to work with the Dual-Tree Complex Wavelet Transform in Häfner et al (2008). The correlation between subbands of different (and equal) color channels is computed based on the mean and standard deviation of coefficient magnitudes.…”

Scalable gastroscopic video summarization via similar-inhibition dictionary selection

“…Many computer-aided endoscopy diagnosis systems have been proposed to assist clinicians in improving the accuracy of medical diagnosis using the images or videos recorded in the inspection of a GT tract. According to the specific lesions, these systems can be classified to handle bleeding [30,31], tumors [32,33], Helicobacter pylori [34], cancer [35,36], Crohn's disease [37] and polyps [38]. Moreover, some other applications include pose detection for endoscopy [39], video segmentation [40] and three-dimensional reconstruction of the digestive wall [41].…”

Deep sparse feature selection for computer aided endoscopy diagnosis

“…Depending on the instruments for gastropathy examination, there are mainly two types: (a) the active flexible gastroscopes, including the traditional endoscopy [7], recent narrow-band imaging (NBI) endoscopy [12], zoom-endoscopy [13,14] and confocal laser endomicroscopy (CLE) [15], which are a thin, flexible fibre-optic instrument passed through the mouth to examine the inside of the gullet, stomach and duodenum; (b) the more recent passive and non-invasive technology, Wireless Capsule Endoscopy (WCE) [2][3][4][5], widely used for small intestine and gullet examination, which captures and sends out the internal images for diagnosis at rate of 2 fps. Depending on the areas in gastrointestinal tract (GI), the methods can be broken down for the esophagus [16], the stomach [17,7], the small intestine [2][3][4][5] and the colon [9,18].…”

“…Depending on the areas in gastrointestinal tract (GI), the methods can be broken down for the esophagus [16], the stomach [17,7], the small intestine [2][3][4][5] and the colon [9,18]. Depending on the specific lesions, the diagnosis methods can be classified to handle bleeding [2], cancer [19,17], Celiac disease, Helicobacter pylori [7], polyps [20,14] and ulcers [4], motility assessment [21], tumors [6,7], Barrett's esophagus, Crohn's disease [9,18], and just classify the region into normal and abnormal [22]. Some other applications include detecting informative frames [3], WCE color video segmentation [23], summarization [24] and clustering [25].…”

“…Theoretically, the computer aided endoscopy diagnosis methods have two key technologies: (1) image representation, in which some color and texture feature units are extracted for endoscopy image analysis, such as Wavelet feature [14], Gabor feature [12], Fourier feature [13], LBP texture [26], various color histograms [27], edge feature [28], invariant feature [29] and feature combination [4,2]. (2) Diagnosis/classification model, in which the state-of-the-art methods [10,11] adopt various models, such as statistical learning models, e.g., Gaussian Mixture Models (GMM); supervised learning models, e.g., SVM [4], Artificial Neural Network (A-NN); the K-Nearest Neighborhood (K-NN) based methods by considering the problem as a retrieval issue.…”