A More Robust Method for Multi-modality Medical Image Registration

Radiological images are increasingly being used in healthcare and medical research. There is, consequently, widespread interest in accurately relating information in the different images for diagnosis, treatment and basic science. The objective of registration process is to obtain a spatial transformation of a floating image to a reference image by which a similarity measure is optimized between the two images. In order to speed up the registration process and avoid local minimum, pre-alignment and Regional Mutual Information has been introduced before. Various strategies of optimization have been tested. And we select bilinear interpolation for all the tests so that the results can be well compared. Experimental results involving clinical images of CT, MR and PET in the brain or chest are presented for rigid registration. Registration with PAX provides well initial transform parameters for the optimizer. And take precision and time into account, Powell optimization is more excellent than others for registration based on RMI.