Quantitative evaluation of fatty liver in ultrasonography

Nowadays, fatty liver is one of the common encountered liver diseases. In addition to imaging techniques such as ultrasound, computed tomography, magnetic resonance imaging, biopsy and blood tests are used in the diagnosis and follow-up of this disease. While technology continues to evolve in various areas, it is still preferable to diagnose diseases with low-cost, easily accessible, fast and non-invasive methods. Ultrasonography is the most commonly used method for the detection and follow-up of liver fat. Ultrasound is a method that is easily accessible, low cost and has no known harm in human health. Although ultrasound imaging method is used frequently to detect fatty liver, there is always a need for an expert opinion to diagnose the disease using ultrasound images. The expert gradates fatty liver by evaluating the findings in the ultrasound images. However, the evaluation of this liver fatty tissue by experts on ultrasound images may cause different interpretations at different times. In other words, the level of fatty liver can be interpreted to be different from the previous one. The aim of this thesis is to propose a mathematical method that can assist the expert and support the expert opinion in objective detection of fatty liver from ultrasound image. Thus, with the quantitative result obtained from the proposed method, the patient can be given more objective information and the expert can be provide convenience to evaluate the ultrasound images at different times.