Automatic Diagnosis of Liver Diseases from Ultrasound Images

Ultrasound is a widely used medical imaging technique. Tissue characterization with ultrasound has become important topic since computer facilities have been available for the analysis of ultrasound signals. Automatic liver tissue characterizations from ultrasonic scans have been long the concern of many researchers. Different techniques has been used ranging from processing the RF signals received by the transducer to using neural networks to analyze images based on image texture. In this paper, an automatic liver diseases diagnostic system is implemented for early detection of liver diseases. The proposed system classification accuracy is 96.125%. The system advantage is its high accuracy and its computation simplicity. The system can be used as a second opinion system to aid the diagnosis of liver diseases