Semiautomatic 3-D image registration as applied to interventional MRI liver cancer treatment

We evaluated semiautomatic, voxel-based registration methods for a new appl ication, the assessment and optimization of interventional magnetic resonan ce imaging (I-MRI) guided thermal ablation of liver cancer. The abdominal i mages acquired on a low-field-strength, open I-MRI system contain noise, mo tion artifacts, and tissue deformation. Dissimilar images can be obtained a s a result of different MRI acquisition techniques and/or changes induced b y treatments. These features challenge a registration algorithm. We evaluat ed one manual and four automated methods on clinical images acquired before treatment, immediately following treatment, and during several follow-up s tudies. Images were T2-weighted, T1-weighted Gd-DTPA enhanced, T1-weighted, and short-inversion-time inversion recovery (STIR), Registration accuracy was estimated from distances between anatomical landmarks. Mutual informati on gave better results than entropy, correlation, and variance of gray-scal e ratio, Preprocessing steps such as masking and an initialization method t hat used two-dimensional (2-D) registration to obtain initial transformatio n estimates were crucial. With proper preprocessing, automatic registration was successful with all image pairs having reasonable image quality. A reg istration accuracy of approximate to 3 mm was achieved with both manual and mutual information methods. Despite motion and deformation in the liver, m utual information registration is sufficiently accurate and robust for usef ul applications in I-MRI thermal ablation therapy.