Early detection of osteoarthritis by 3D computed tomography

Abstract

Osteoarthritis is a common disease that constitutes a considerable socioeconomic burden. So far no radiological procedure was available that allowed the disease to be detected in an early stage. Monitoring disease progression and efficacy of therapeutic procedures was also not possible.

We present a measuring procedure and describe evaluation tools that are based on 3D computed tomography. The procedure provides quantitative data on pathological features known to be associated with osteoarthritis such as subchondral sclerosis, osteophytes, joint space narrowing, and bone lesions. It is fully three dimensional, the results are operator independent and describe global as well as local parameters. Changes in these parameters are evaluated from differential images. For that purpose a 3D matching procedure based on translation and rotation-specific correlation techniques is used that allows a repositioning of the bone objects with a mismatch of less than 0.3 mm with regard to translations and 0.250 with regard to rotations.

The use of the new tools for measurement, repositioning and evaluation is illustrated with the results from a clinical study. Patients with mild osteoarthritis of the knee joint, healthy controls and osteoporotic patients, all in the same age range of 60 to 75 years, were examined during three years. The cortical to trabecular bone ratio describing subchondral sclerosis proved to be especially helpful to distinguish early osteoarthritic cases from age matched controls. Disease progression is best shown by the increase in the number of detected bone lesions, a reduction of the mean distance between femur and tibia and an increase in the cortical to trabecular bone ratio. Hence the presented procedure appears to have a great potential in diagnosis and treatment monitoring of osteoarthritis.