Abstract:
The classical Bayes rule plays very important role in the field of lesion identification. However, the Bayesian approach is very difficult in high dimensional spaces for lesion detection. An alternative approach is Principle Component Analysis (PCA) for automatic multiple sclerosis lesion detection problems in high dimensional spaces. In this study, PCA based Bayesian approach is explained for automatic multiple sclerosis lesion detection using Markov Random Fields (MRF)and Singular Value Decomposition (SVD). It is shown that PCA approach provides better understanding of data. Although Bayesian approach gives effective results, itis not easy to use in high dimensional spaces. Therefore, PCA based Bayesian detection will give much more accurate results for automatic multiple sclerosis (MS)lesion detection.
