Comprehensive Summary
This study compared the performance of coronal fast low-angle shot (FLASH) and sagittal double echo steady state (DESS) MRI sequences for detecting longitudinal cartilage thickness changes in knee osteoarthritis using a fully automated AI-based segmentation method. The dataset analyzed included subjects from the FNIH biomarker cohort with baseline and two-year follow-up MR images. The results demonstrated strong cross-sectional agreement between the two MRI sequences, but longitudinal changes in cartilage thickness were more sensitively detected by the sagittal DESS sequence, particularly in the medial tibiofemoral compartment. Automated segmentation performed with high reproducibility, allowing robust quantification of small cartilage changes over time.
Outcomes and Implications
These findings validate the use of fully automated segmentation for longitudinal cartilage thickness assessment and highlight the superior sensitivity of sagittal DESS MRI for monitoring osteoarthritis progression. Such advances in automated MRI analysis enhance the efficiency and precision of OA biomarker studies, facilitating early detection of disease modification in clinical trials.