Intelligent systems using fuzzy logic for the determination of breast tumors

Describes an intelligent system using fuzzy logic for the determination of breast tumors in ultrasonic images. The system is based on the fuzzy logic and relaxation method for determining a surface region of breast tumor using ultrasonic voxel data. A significant feature of the system is that the membership functions are approximated by Rician density function and their parameter is determined by the histogram of the region that is classified by three dimensional LoG (Laplace of Gaussian) filtering. The authors applied the proposed algorithm to a total of seventy clinical cases, forty malignant and thirty benign tumors. As a result, they obtained a successful extraction and a good agreement with the evaluation of tumors using a conventional diagnosis