PRINCIPAL AXES AND SURFACE FITTING METHODS FOR 3-DIMENSIONAL IMAGE REGISTRATION

We evaluated the effect of the image acquisition parameters on the acc uracy of the principal axes and surface-fitting techniques for three-d imensional image registration. Using two types of phantom objects, MR brain image and a mathematically defined ellipsoid, we simulated pairs of scans with known acquisition parameters, including longitudinal co verage, magnitude of misregistration, number of sections and section t hickness. Both methods are sensitive to the systematic deformation of contours. The principal axes method is also sensitive to incomplete sc an coverage and to the x-axis and y-axis misangulation. Both methods a re insensitive to the number of sections, section thickness and the nu mber of points per section. Surface fitting performed well without use r supervision. There is no need for routine inclusion of the scaling f actors as search parameters. The results confirm the feasibility of th ree-dimensional multimodality registration of brain scans with accurac y 1-2 mm, with surface fitting being the method of choice.