Application of three-dimensional magnetic resonance image registration for monitoring hip joint diseases Masaki Takao a, T , Nobuhiko Sugano a , Takashi Nishii a , Hisahi Tanaka b , Jun Masumoto c , Hidenobu Miki a , Yoshinobu Sato c , Shinichi Tamura c , Hideki Yoshikawa a a Department of Orthopaedic Surgery, Osaka University Graduate School of Medicine, Osaka 565-0871, Japan b Department of Radiology, Osaka University Graduate School of Medicine, Osaka 565-0871, Japan c Division of Interdisciplinary Image Analysis, Department of Medical Robotics and Image Sciences, Osaka University Graduate School of Medicine, Osaka 565-0871, Japan Received 28 July 2004; accepted 3 February 2005 Abstract The purpose of this study was to estimate the accuracy of a method in which three-dimensional (3D) magnetic resonance (MR) volume registration is used for monitoring hip joint disease. Data were analyzed using a normalized cross-correlation (NCC) algorithm involving a user-selected 3D box including the proximal femur. Most of the femoral head was not included in the 3D box because it can become deformed during the course of disease. The accuracy of registration around the femoral head was evaluated using five phantoms and clinical MR data of 17 patients with hip joint disease. In the phantom experiment, registration accuracy was evaluated using four fiducial markers attached to the femoral head. In the experiment using clinical data, registration accuracy was evaluated using a landmark in the femoral head. The registration accuracy in the phantom and clinical experiment was 0.43F0.18 mm (S.D.) and 1.12F0.46 mm (S.D.), respectively. The former is a value less than half the minimum dimension of a voxel (1.251.251.0 mm). Although the latter is slightly larger than the minimum dimension of a voxel, actual errors would be smaller because of the uncertainty in landmark localization. In conclusion, the present method based on an NCC algorithm can be used to accurately register serial MR images of the femoral heads with an error on the order of a voxel. We believe that this method is sufficiently accurate for monitoring hip joint diseases. D 2005 Elsevier Inc. All rights reserved. Keywords: Hip joint disease; Monitoring; Magnetic resonance imaging; Volume registration; Normalized cross-correlation 1. Introduction In clinical practice, precise evaluation of lesion changes on serial medical images is very important for monitoring disease activity and assessing response to therapy. However, it is impossible to obtain multiple medical images of the pa- tient in the same position at different times. This repositioning error in targeting of organs or tissue between scans affects the accuracy of diagnosis and treatment assessment. To resolve this problem, image registration techniques have been developed. Image registration is a computational process that realigns images of the same subject acquired at different times [1]. Various types of image registration techniques have been developed to improve the accuracy of matching serial images [1]. A volume registration algorithm that realigns serial images based on voxel values is widely used for regis- tration of serial MR images of brain diseases such as brain tumors [2], multiple sclerosis [3,4], Alzheimer’s disease [5] and schizophrenia [6]. There have been few studies of the application of image registration to monitor musculoskeletal disease. In patients with joint disease, positional matches between image sets are usually poor because most patients have pain or contracture in the affected joint or neighboring joints. It would be very useful to improve the alignment of such image sets using image registration. The hip joint is one of the main sites affected by osteoarthritis, osteonecrosis, rheumatoid arthritis, amyloid arthritis and rapidly destructive coxopathy. Reports indicate that magnetic resonance imaging (MRI) is a sensitive imaging modality for monitoring these diseases [7–15]. 0730-725X/$ – see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.mri.2005.02.002 T Corresponding author. Tel.: +81 6 6879 3552; fax: +81 6 6879 3559. E-mail address: masaki-tko@umin.ac.jp (M. Takao). Magnetic Resonance Imaging 23 (2005) 665 – 670