Evaluation of Optimization Methods for Nonrigid Medical Image Registration Using Mutual Information and B-Splines

• Published in: IEEE transactions on image processing 2007-12, Vol.16 (12), p.2879-2890
• Main Authors: Klein, S., Staring, M., Pluim, J.P.W.
• Format: Article
• Language: English
• Subjects: Algorithms; Applied sciences; B-splines; Biological and medical sciences; Biomedical imaging; Computation; Computed tomography; Computerized, statistical medical data processing and models in biomedicine; Convergence; Cost function; Deformation; Derivatives; Exact sciences and technology; Humans; Image Enhancement - methods; Image Interpretation, Computer-Assisted - methods; Image processing; Image registration; Information, signal and communications theory; Iterative algorithms; Iterative methods; Medical; Medical management aid. Diagnosis aid; Medical sciences; Mutual information; Nonlinearity; nonrigid image registration; Numerical Analysis, Computer-Assisted; Operations research; Optimization; Optimization methods; Pattern Recognition, Automated - methods; Reproducibility of Results; Sensitivity and Specificity; Signal processing; Spline; Studies; subsampling; Subtraction Technique; Telecommunications and information theory; Testing
• ISSN: 1057-7149; 1941-0042
• DOI: 10.1109/TIP.2007.909412

A popular technique for nonrigid registration of medical images is based on the maximization of their mutual information, in combination with a deformation field parameterized by cubic B-splines. The coordinate mapping that relates the two images is found using an iterative optimization procedure. This work compares the performance of eight optimization methods: gradient descent (with two different step size selection algorithms), quasi-Newton, nonlinear conjugate gradient, Kiefer-Wolfowitz, simultaneous perturbation, Robbins-Monro, and evolution strategy. Special attention is paid to computation time reduction by using fewer voxels to calculate the cost function and its derivatives. The optimization methods are tested on manually deformed CT images of the heart, on follow-up CT chest scans, and on MR scans of the prostate acquired using a BFFE, Tl, and T2 protocol. Registration accuracy is assessed by computing the overlap of segmented edges. Precision and convergence properties are studied by comparing deformation fields. The results show that the Robbins-Monro method is the best choice in most applications. With this approach, the computation time per iteration can be lowered approximately 500 times without affecting the rate of convergence by using a small subset of the image, randomly selected in every iteration, to compute the derivative of the mutual information. From the other methods the quasi-Newton and the nonlinear conjugate gradient method achieve a slightly higher precision, at the price of larger computation times.