Assessment Method for Tumorous Objects Volume from Brain MR Images

摘要

The work is devoted to describe our automated technique of tumour and neoplasm volume assessment on brain MRI.First we proposed adaptive processing of dynamic local surroundings by estimating of local entropy. Secondly, theinvestigation of volume assessments directly is ground on the proposition to check extent of interdependency betweenadjacent slices. The overall approach uses double-criterion principle. The first one estimates cross correlation, the secondone is based on Max-dependency principle. Obtained results prove consistency of the implemented approaches.