Predictive factors of osteoarthritis progression are not yet well understood. However, a growing role attaches importance to the subchondral bone. The aim of the present project is to determine predictive factors of progression of osteoarthritis at the knee by a multimodal characterization of subchondral bone by Medical Resonnance Imaging, direct high resolution digitization radiographs and bone texture analysis. At the end of the project, an innovative imaging device, combining semi-automatic softwares for texture analysis, control detection and image registration would be supplied. This will enable on the one hand a more accurate and reproducible way to measure the joint space width of the affected compartment and on the other hand, an assistance to better detect patients at risk of progression of their knee osteoarthritis. Identifying These "progressors" patients might permit their selection in clinical trials at baseline adapted to their severe disease, using for example biologic treatments targeting knee osteoarthritis. The main objective of this study is to analyze the predictive capacities of bone texture parameters measured on the high-resolution radiography of the knee on the structural evolution of the knee osteoarthritis at 3 years.

Inclusion Criteria:

• knee osteoarthritis according to the criteria of knee osteoarthritis of American College of Rheumatology.
• Having felt a painful push in the last 6 months.
• Visual Analogic Scale pain > 4 on a 0 to 10 scale.
• having a Radiographic score of OARSI (Osteoarthritis Research Society International Score) from 2 to 3 whether it is for the presence of a joint space narrowing or the presence of an osteophyte

Exclusion Criteria:

• Patello-femoral degenerative osteoarthritis isolated.
• Incapacitated to consider the position in schuss for the realization of the radiography.
• Valgus > in 5 °
• Secondary degenerative osteoarthritis