Local Area Cartilage Segmentation: A Semiautomated Novel Method of Measuring Cartilage Loss in Knee Osteoarthritis

Abstract

OBJECTIVE To assess the responsiveness and reader time of a novel semi-automated tool to detect knee cartilage loss over two years in subjects with knee OA. METHODS 122 subjects from the OAI Progression Cohort were selected. A reader used the software method to segment cartilage on DESS scans in the medial compartment of the femur from the baseline and 24-month visits. Change in cartilage volume (ΔV) was measured at a fixed weight-bearing (WB) location with respect to the three-dimensional coordinate system based on cylindrical coordinates. Change was measured for five regions of varying WB surface area centered on the fixed point. The average change (ΔV), the standard deviation (SD) of ΔV and the standardized response mean (SRM) are reported. RESULTS The SRM value was −0.52 for the largest region and decreased in magnitude as smaller regions of cartilage were probed. The average evaluation time was less than 20 minutes per knee compartment, split approximately evenly between a technician and a trained reader. CONCLUSION The results establish that measurement of cartilage loss in a local region can be done efficiently and that the resultant measures are responsive to loss of cartilage over time. The coordinate system can potentially be used to objectively examine and establish a consistent location for all knees that is most responsive to change in cartilage volume. This technique can provide an objective quantitative measure of cartilage loss rapidly and could substantially reduce study costs for large trials and datasets.

Cluster analysis of quantitative MRI T 2 and T 1ρ relaxation times of cartilage identifies differences between healthy and ACL-injured individuals at 3T

“…This approach was shown to be more sensitive to local patterns in T 1ρ relaxation time than traditional ROI-based approaches. Another potential approach, which has been used to study variations in 3D cartilage thickness 34,35 , is to use projection maps; these maps could also be used for registration between time-points and for isolating heterogeneity and focal defects. Capturing as much information about pathological changes and improving the detection of these changes is very important for identifying early cartilage degeneration and monitoring therapy.…”

Quantitative measurement of medial femoral knee cartilage volume – analysis of the OA Biomarkers Consortium FNIH Study cohort

“…All knees were evaluated at a fixed measurement location in the central weight-bearing portion of the medial femur as described in a previous publication [10]. Briefly, the measurement region was based on two axes of a cylindrical coordinate system, z and θ. z roughly corresponded to the medial-lateral direction (greater z was more medial) and θ to the anterior-posterior part of the articular surface of the femur (greater θ was more posterior).…”

“…The LACS method was previously validated for intra-reader and inter-reader reliability and showed good reproducibility with intraclass correlation coefficients (ICC) > 0.9) [10]. In a recent published study the repositioning error of the LACS method for Δz = 0.1 was measured on 10 healthy volunteers with a modified DESS-sequence twice on the same day where the subjects were removed and placed back into the MRI scanner in-between scans.…”

“…We previously developed and validated an efficient, reproducible, and responsive quantitative software tool to measure cartilage volume in focal locations on the medial femur on MRI scans [10, 11]. The local-area cartilage segmentation (LACS) method uses anatomical landmarks and a mathematically robust coordinate system to identify consistent regions of cartilage for fast segmentation.…”

Dual-pathway multi-echo sequence for simultaneous frequency and T2 mapping

“…The protocol was identical to that recommended by the Osteoarthritis Initiative (41, 42): water-only excitation, sagittal orientation, flip angle 25°, resolution = 0.37×0.46×0.7 mm 3 , TR = 16.3 ms, TE + / TE − = 5.3 / 10.8 ms (referenced DESS). For our multi-TE approach, TE + = 3.35 / 9.03 / 10.73 ms and $T{E}_j^{-}$ = 5.34 / 7.04 / 12.72 ms. Cartilage segmentation was performed for both datasets as in (43). …”