Ultrasound Chest Muscle Characterization using 2D Texture Analysis

Investigation and characterization of a musculoskeletal organ can be performed using ultrasound imaging. The images can be examined in term of morphological and textural analysis. Statistical texture analysis is one of the established techniques for image analysis and classification. It is suitable for investigation and characterization of the musculoskeletal ultrasound image. In this study, the statistical texture analysis was performed for characterization of Pectoralis minor (Pmi) and Pectoralis major (Pma) muscle of the chest. The images were captured using ultrasound machine and the region of interest (ROI) was obtained for Pma and Pmi muscle area. The ROI images were analyzed using First-order and Second-order statistical texture analysis in MATLAB software. The statistical texture feature values of Pmi and Pma were obtained and compared. The result shows that the First-order texture features of variance and kurtosis and the Second-order texture features of contrast, correlation, energy, and homogeneity tend to provide different characteristics in Pmi and Pma muscle image classification.