Evaluation of multimodal medical image registration based on Particle Filter

This paper presents a performance evaluation of a new multimodal image registration algorithm which is based on Bayesian estimation theory, specifically on Particle Filters. The results point to an efficient, easy to implement and robust to noise algorithm. The registration method showed good performance when using partial data, and it was compared with an algorithm based on maximization of mutual information and a Hyperplanes optimization method. Finally, we showed that the algorithm may be parallelizable, so that it is possible to reduce the computation time for image registration. © 2010 IEEE.

Bayesian estimation; Image registration; Multimodal images; Particle filters Bayesian estimation theory; Bayesian estimations; Computation time; Multimodal image registration; Multimodal images; Multimodal medical images; Mutual informations; Optimization method; Partial data; Particle filter; Performance evaluation; Registration methods; Algorithms; Automation; Bayesian networks; Control; Distributed computer systems; Electrical engineering; Estimation; Medical imaging; Nonlinear filtering; Optimization; Process control; Target tracking; Image registration

VIVO