Medical ultrasound imaging is widely used for the diagnosis of various diseases because of its noninvasiveness and real-time performance. However, due to the ultrasound-specific phenomena, it is difficult to understand what is visualized on ultrasound images without the knowledge and experience of the examiner. To overcome this difficulty, many quantification techniques have been developed. This presentation will outline the factors that make ultrasound images difficult to understand. I will then describe our recent work to quantify and visualize liver diseases based on the statistics-based analysis of ultrasound echo signals.