Variable fiber orientations of knee cartilages investigated by zonal T2* measurements with automatic segmentation

Osteoarthritis (OA) is a common disease in aged people, which results in degeneration of the articular cartilage and the meniscus [1]. Identification of the tiny changes of these cartilages is important to the early detection of OA and the monitoring of the disease progression. Recently, quantitative MR measurements, such as mapping of T1ρ, T2, and diffusion coefficient, have been used to evaluate the architecture of the knee cartilages [2]. However, in the thin layer of articular cartilage, manual selection of the regions-of-interest may become the major source of discrepancy for zonal analysis of the collagen fiber orientations with different extents of magic angle effect [3]. Therefore, the purpose of this study is to propose an efficient image segmentation method based on the 2D fussy C-means (FCM) algorithm [4] to facilitate MR T2* measurements, and to investigate the zonal difference of knee cartilages at variable fiber orientations.