Comparison of the manual and computer-aided techniques for evaluation of wrist synovitis using dynamic contrast-enhanced MRI on a dedicated scanner

Abstract

Objective: Traditional methods for assessment of synovial inflammation in rheumatoid arthritis such as clinical examination, immunohistology of bioptic samples, scintigraphy, and radiography have several limitations, including lack of sensitivity, need of invasive techniques, and administration of radioactive material. MRI lacks on standardisation and the data are often analysed using laborious, relatively rigid scoring methods.

Materials and methods: This study introduces a standardized computer-aided method for quantitative analysis of MRI of the wrist on a dedicated scanner. Assessment of the synovial inflammation was performed using a semi-automated model-based method in conjunction with patient motion reduction algorithms. Further, the new method was compared with the traditional user-dependent ROI-based technique.

Results: The computer-aided technique generated robust and reproducible results. Application of motion reduction algorithms allowed for significant improvements of the signal to noise ratio, which is especially important in the datasets acquired with low-field scanners.

Conclusion: The use of the computer software can be beneficial for diagnostic decision in cross sectional as well as longitudinal MRI examinations of the wrist in rheumatoid arthritis.