A texture classification method for diffused liver diseases using Gabor wavelets

Abstract

We proposed an efficient method for classification of diffused liver diseases based on Gabor wavelet. It is well known that Gabor wavelets attain maximum joint space-frequency resolution which is highly significant in the process of texture extraction and presentation. This property has been explored here as the proposed method outperforms the classification rate obtained by using dyadic wavelets and methods based on statistical properties of textures. The feature vector is relatively small compared to other methods. This has a significant impact on the speed of retrieval process. In addition, the proposed algorithm is not sensitive to shift of the image contents. Since shifting the contents of an image will cause a circular shift of the Gabor filter coefficients in each sub-band. The proposed algorithm applied to discriminate ultrasonic liver images into three disease states that are normal liver, liver hepatitis and cirrhosis. In our experiment 45 liver sample images from each three disease states which already proven by needle biopsy were used. We achieved the sensitivity 85% in the distinction between normal and hepatitis liver images and 86% in the distinction between normal and cirrhosis liver images. Based on our experiments, the Gabor wavelet is more appropriate than dyadic wavelets and statistical based methods for texture classification as it leads to higher classification accuracy.

Core samples for radiomics features that are insensitive to tumor segmentation: method and pilot study using CT images of hepatocellular carcinoma

“…However, our study showed that in HCC, and most likely, in many other tumors, there is high variability in human determination of tumor outlines, leading to increased uncertainty in the features derived from the borders. However, the features that have shown high robustness (Table 5) have been reported to be correlated with clinical variables in previous studies (Gabor filters, 39,40 Daubechies, 41 and Haralick features 42,43 ). Future studies are needed to further investigate the association of these features with clinical variables.…”

Usefulness of textural analysis as a tool for noninvasive liver fibrosis staging

“…The resolution cell is assumed to be 0.28 9 0.28 mm 2 . Computation times are evaluated for a single core algorithms and the papers that use these algorithms in fibrosis detection (when possible): first order statistics [13,34], gray tone difference matrix [33], gray level cooccurrence matrix [7,13,34,35], multiresolution fractal dimension [7], differential box counting [8,36], morphological fractal dimension estimators [37], Fourier power spectrum [6,7], Gabor filters [12], Law's energy measures [7], texture edge co-occurrence matrix [8], phase congruency [24], and texture feature coding matrix [14]. Twelve algorithms that process the entire ROI were implemented and 234 features were computed per image.…”

“…There are several approaches found in the literature. Some authors have evaluated the RF signal [1,2], while others have proposed some visual scores [3,4], but most authors have employed a texture analysis system to discriminate between various fibrosis stages [5][6][7][8][9][10][11][12][13][14].…”

Analysis and Comparison of Color Features for Content Based Image Retrieval

“…Table 1 showsresults of the experiments, which were carried out based on the weak relevance judgments. The majority of known CBIR systems, such as VisualSEEk [17], QBIC [18], Mars [19,20], and Netra [21], use color histograms in retrieval by color. DCD color methods similarity matching does not fit human perception very well, and it will cause incorrect ranks for images with similar color distributionThe above allows us to assume that, under appropriate selection of the color space, its quantization, and metrics on the feature space, DCD color can be quite effective in retrieval by color.…”