Discovering the cirrhosis grades from ultrasound images by using textural features and clustering methods

Abstract

Cirrhosis characterization and grading is an important issue nowadays in the medical domain, as this disease can lead to death. We aim to discover the cirrhosis grades in a noninvasive manner, using computerized methods. Concerning the feature computation, we chose the texture-based methods, as they revealed subtle aspects of the tissue, not detectable by the human eye. For this purpose, we used first, second and third order statistics of the gray levels, edge-based statistics, statistics of the textural microstructures, and also textural features computed at multiple resolutions, after applying the Wavelet transform. All these features were inputs to clustering methods, such as k-means clustering and expectation maximization (EM), implemented for the determination of the cirrhosis grades, each grade corresponding to a certain cluster. The relevant textural features, for each discovered grade, were also identified, by computing a specific score, for each feature, based on the result of the clustering methods.

Diseased tissue area detection and delimitation, by fusion between finite difference methods and textural analysis

“…Supervised classifiers were applied, in order to validate the textural model and to perform automatic diagnosis, using methods such as the Decision Trees, Artificial Neural Networks, Support Vector Machines, as well as classifier combinations, using the bagging, boosting and stacking schemes [10]. We also performed unsupervised classification, by using clustering methods, in order to determine the severity grades of the diseases [12].…”

“…These features are usually: GLCM Homogeneity, GLCM Correlation, GLCM Contrast, Autocorrelation Index, Edge Orientation Variability, Spot Frequency, Wavelet Frequency, Haar GLCM Entropy, Haar GLCM Variance of order 3, Haar GLCM Homogeneity of order 3. [9], [12] In such a way, the quite imprecise contour of the investigated diseased tissue zone, obtained by iteratively identifying the limits of similar texture characterized areas, is consolidated by means of the final position of the CrankNicolson or Explicit Finite Difference method based active contour. The results of our experiments are presented in the following table (Table I.…”

“…vector-valued algebraic equation, which approximates the ELP-Equation (9), is obtained from (9), (11), (12) and (13), as follows:…”