Magnetic Resonance Imaging Based Evaluation of Articular Cartilage Degeneration

Introduction: In-vivo magnetic resonance imaging(MRI) can potentially be used to determine structural changes in articular cartilage of the knee. Water, the dominant component in articular cartilage, is present in both free(70%) and bound form. Bound water is present in collagen fibrils comprising 1020% of the cartilage. Studies indicate that T2 – the proton spin-spin relaxation time parameter correlates with histologically observed variation arising from structural anisotropy of the collagen fibrillar network(1). Normally, the fibrils are oriented in response to mechanical loading forces of the knee joint, but with progression of osteoarthritis(OA), structural disruption occurs. This breakdown results in fluctuations in the ratio of free to bound water in the cartilage and is manifested in T2 variation. There are few cross-sectional studies of T2 changes with OA(2-3), especially in humans. This study evaluates T2 differences and their potential relationship with cartilage morphometry in healthy volunteers and in patients in varying stages of OA, as ascertained by the radiography based Kellgren Lawrence(KL) scale. Materials/Methods: In-vivo sagittal MR images of the knee(n=21; normal=7, mild radiographic OA1(KL1-2)=7, severe radiographic OA2(KL3-4)=7) are acquired on a 1.5 tesla GE Signa system. A USA Instruments phased array coil is used for imaging. The imaging protocol combines a high resolution (0.234x0.234x2mm) 3D SPGR with a 2D dual spin echo(TE1=10ms; TE2=45ms,TR=1500ms), T2-weighted(0.468x0.468x4mm) sequence, both with fat suppression. The sequences are optimized for accurate cartilage segmentation and generation of T2 maps across the various anatomical cartilage compartments(tibial, femoral and patellar), respectively. Images are obtained over identical locations, thereby providing complementary information about morphometric(volume, thickness) and structural(T2) variation associated with anomalous loading of the knee joint. The high resolution image volume is used for segmentation of the cartilage from the surrounding tissue using the immersion based watershed algorithm(4). Apart from morphometric measurements the segmented cartilage volume is registered to the T2 weighted image volume to define the cartilaginous region of interest(ROI) for computation of T2 maps. Intensity variation with cartilage structure cannot be visually observed from in-vivo images, at the current resolution and gradient strength. However, T2 maps computed reflect T2 variation across the patellar, femoral and tibial cartilage. These maps are generated, on a pixel by pixel basis, by extrapolating from the signals obtained at two different echo times(TE1,TE2). The MR signal is given by eqn. [1] and T2 derivation is shown in [2]-[3]. SI ∝N( H)(1− −TR/ T1 e ) −TE / T2 e [1],