Texture analysis as a noninvasive tool for fibrosis assessment in chronic hepatitis C. influence of expert dependent variability over the performance of texture analysis

Texture analysis is viewed as a method to enhance the diagnosis power of classical B-mode ultrasound image. Present study aims to evaluate the dependence between the human expert and the performance of such a texture analysis system in predicting the cirrhosis in chronic hepatitis C patients. 125 consecutive chronic hepatitis C patients were included in this study. All the patients had positive HCV-RNA in serum and had undergone percutaneous liver biopsy for disease staging using Metavir score. Ultrasound images were acquired from each patient and 4 experts established regions of interest. Textural analysis software generated 234 features from each region of interest. Relevant textural features were identified and a classification schema was evaluated. Texture analysis can discriminate between F0 and F4 fibrosis stages (AUROC=0.64). The performance of this approach depends highly on the human expert that establishes the regions of interest (p<;0.05). The relevant textural features were identified and it was shown that the detection performance didn´t depend on the particular feature selection (p=0.8). In classical form met in literature non invasive diagnosis through texture analysis has limited utility in clinical practice because of the user variability introduced by the expert who establishes the regions of interest.