Nonparametric method for detection and identification of regions of concern in multidimensional intensity images

摘要

A method for detection and identification of regions of concern in multidimensional intensity images using a template and a nonparametric test is presented. Data corresponding to background intensity is extracted from the outer region of the template while data for mass detection is extracted from the inner region. The statistical nature of the background intensity, specifically the tail structure of the unknown distribution, is estimated via tolerance intervals. Mass detection is based on the fraction of test data within the tail structure.