Non-rigid Multimodal Image Registration Based on the Expectation-Maximization Algorithm

Abstract

Abstract. In this paper, we present a novel methodology for multimodal non-rigid image registration. The proposed approach is formulated by using the Expectation-Maximization (EM) technique in order to estimate a displacement vector field that aligns the images to register. In this approach, the image alignment relies on hidden stochastic random variables which allow to compare the intensity values between images of different modality. The methodology is basically composed of two steps: first, we provide an initial estimation of the the global deformation vector field by using a rigid registration technique based on particle filtering, obtaining, at the same time, an initial estimation of the joint conditional intensity distribution of the registered images; second, we approximate the remaining deformations by applying an iterative EM-technique approach, where at each step, a new estimation of the joint conditional intensity distribution and the displacement vector field are computed. The proposed algorithm was tested with different kinds of medical images; preliminary results show that the methodology is a good alternative for non-rigid multimodal registration.

Multimodal image registration based on the expectation–maximisation methodology

“…Hence, an advantage of our methodology in terms of accuracy and robustness is visualised in the synthetic and practical scenarios for the parametric registration, and the elastic version. In fact, this work expands our initial contribution in [29] by four major aspects: an updated version of the EM formulation and its analytical derivation; a new parametric strategy for image registration; a comprehensive evaluation of both registration methodologies, parametric and elastic; and an updated revision of the state of the art.…”

Effect of optimization framework on rigid and non-rigid multimodal image registration

“…They used structural similarity to optimize the framework. Arce-Santana et al [13] made a new approach towards multimodal image registration when they used expectation-maximization (EM) to calculate displacement vectors that are used in non-rigid image registration algorithm to form the transformation matrix. In 2015, Vicente et al [2] discussed about a 3D registration technique which involved multimodal images of anatomic (MRI) and functional (fMRI and PET) brain data.…”

“…But in this article, a comparative study of multimodal image registration is presented, and its performance optimization based on ant colony optimization and flower pollination algorithms [11] is analyzed. Additionally, the proposed image registration framework includes both rigid and non-rigid registration algorithms [12,13].…”

“…Although we only investigated these two finest metaheuristic algorithms [12] on monomodal images on the Demons registration framework, we chose also to use it in the proposed framework. Previously, not much has been done to optimize the performance of the multimodal [13,14] registration. Hence, the current work aims to present an in-depth analysis of the performance optimization of multimodal registration framework, so that the framework can work better than the existing multimodal registration models.…”