A Triple Level Registration Strategy of Medical Image Based on Mutual Information / 航天医学与医学工程

ABSTRACT

Objective To further solve the mutual restricted problems among precision, speed and disregistrated rate of registration algorithm based on mutual information (MI). Methods A new strategy on the search of registration parameters, a triple level registration strategy, was proposed. It searched the translation parameters and the rotation parameters respectively. Based on analyzing the advantage and limitation of MI-based registration, Euclid closeness of Fuzzy theory was introduced into the process of the triple level registration. Results The computer simulation results demonstrated that the proposed registration method, on the condition of retaining the same precision, improved both the speed (step numbers and registration time) and the disregistrated rate. Conclusion The proposed triple level registration strategy is a registration strategy to be able to harmonize precision, speed and disregistrated rate effectively.