Breast mass Classification with shape and texture based features in processing ultrasound image

Tissue differentiation methods and tools with the help of computer- assisted diagnosis based on physical measurement and help ultrasound beam , the target is ideal. Seeking to achieve this goal , a powerful processing system and a comprehensive and reliable database of ultrasound images was necessary , due to being academically and provide reliable results , the software Matlab computational used. Finally, a comprehensive system using image processing areas of texture , shape and mass for the diagnosis of breast cancer in ultrasound images were designed and built . Using this method the number of unnecessary surgical expenses decreased to 90% of patients . In terms ofaccuracy, sensitivity and acceptable ultrasound images are approved , which means it can be reasonably sure of the ultrasound images with the least amount of time lump in the breast tissue detected . As a result of experiments conducted in this way, according to 85 % of its mass and is properly diagnosed .

Keywords:

breast cancer , image processing ultrasound images