The Osteoarthritis Initiative (OAI) is a national, multi-center, large cohort, natural history and prevalence study in knee OA that is an effort to set up a public database of images in -5000 subjects using 4 dedicated MR scanners (3 Tesla). The database in addition to clinical data, serum, etc. will contain radiographs for grading of joint space and impact of alignment on knee OA, MR images that permit radiological grading of OA from MR images at 3T, cartilage volume and T2 estimation. The data will be available to the community for analysis. In OA, matching identical regions of interest, such that the exact same cartilage compartment (such as medial femur, lateral femur etc.) can be evaluated longitudinally is of utmost importance. In this U01 application we propose specific developments that will characterize the spatial heterogeneity of T2 and ensure that comparisons of MR derived cartilage morphology and T2 characteristics are made (i) in identical regions between subjects at baseline and (ii) in identical regions in longitudinal follow-up scans for the same subject. Using a subset of images from the current data release (200 images of the knee) and the proposed release of 160 images from the OAI progression cohort, with close interactions with a scientific advisory board, we will develop (i) a shape-based atlas of the knee, (ii) a knee registration algorithm which will enable us to register all baseline and follow-up scans to the knee atlas, so that all scans are analyzed in a common coordinate system with common shape based anatomical regions, (iii) define anatomical compartments such as the medial and lateral femur, patello-femoral compartment, medial and lateral tibia, and determine cartilage thickness and volume in those compartments, (iv) develop a graphical user-interface and platform for the software above for use by other researchers. We will also use the OAI images to generate T2 maps and use texture measures such as entropy to obtain measures of T2 spatial heterogeneity in the compartments. Our ultimate aim is to use the tools developed to assess longitudinal changes in cartilage morphology, T2 and determine whether baseline T2 and its heterogeneity predicts progression, as determined by changes in cartilage morphology, using a subset of images from the progression cohort. The tools developed will then be disseminate to one or two interested investigators for further testing.
