A new quantitative 3D approach to imaging of structural joint disease.

Imaging of joints with 2D radiography has not been able to detect therapeutic success in research trials while 3D imaging, used regularly in the clinic, has not been approved for this purpose. We present a new 3D approach to this challenge called joint space mapping (JSM) that measures joint space width in 3D from standard clinical computed tomography (CT) data, demonstrating its analysis steps, technical validation, and reproducibility. Using high resolution peripheral quantitative CT as gold standard, we show a marginal over-estimation in accuracy of +0.13 mm and precision of ±0.32 mm. Inter-operator reproducibility bias was near-zero at -0.03 mm with limits of agreement ±0.29 mm and a root mean square coefficient of variation 7.5%. In a technical advance, we present results from across the hip joint in 3D with optimum validation and reproducibility metrics shown at inner joint regions. We also show JSM versatility using different imaging data sets and discuss potential applications. This 3D mapping approach provides information with greater sensitivity than reported for current radiographic methods that could result in improved patient stratification and treatment monitoring