Two-dimensional Fractional Brownian Motion in the analysis of medical images

This paper presents the preliminary results of the application of a two-dimensional Fractional Brownian Motion ( FBM ) model of image distribution. This model has proven successful in the interpretation of linear data, as well as the segmentation of images. The approach taken by the authors differs from previous work in that rather than extracting linear data from the image and using a linear FBM model, the actual two-dimensional distribution is employed. It is found that the resulting segmentation can be much more accurate than previous methods. However, due to the computational intensity of the problem, the algorithm as presently encoded results in a binary segmentation, rather than a 1:1 mapping from fractal dimension to grey level. Regardless of the particular type of segmentation produced, the sensitivity of the segmentation in the test cases exhibits great promise for this approach.