Integrating semi-automated radiographic analysis and biomarkers for detecting knee osteoarthritis: cross-sectional study in Central Java, Indonesia
摘要
Radiographic determination of Kellgren and Lawrence (KL) in knee osteoarthritis (OA) has limitation in providing accurate and consistent results. Therefore, improved methods utilizing the automated measurement and biomarkers in detecting knee OA in the early stages is necessary.
ObjectivesThis study examines the differences in the results between semi-automatic and manual measurement and correlation of biomarkers (IL-6, COMP, and CTX-II) to the severity of knee osteoarthritis.
MethodsThe research employed a cross-sectional study involving 126 patients, for whom semi-automated radiographs of the right and left AP weight-bearing knee projections were performed and evaluated using semi-automatic and manual assessment. A digital assessment used MATLAB software to calculate the area and compared it with the manual. Biomarkers were obtained from blood serum for COMP and IL-6, and from urine for CTX-II. Differences between groups were analyzed using Wilcoxon or T-tests, while the Spearman correlation test and linear regression test examined the relationship between the severities of OA based on KL Grades and biomarkers.
ResultsThe area of osteophyte and TFA was no different (p > 0.005). (p < 0.001) between manual and semi-automated measurement while especially on JSW and eminence assessment showed significant different. The grade I and II knee OA threshold is a medial JSW = 4.95 mm; Lateral JSW = 5.35 mm; osteophyte area = 10.6 mm2; eminence = 8.75 mm. The higher levels of COMP, CTX-II, and IL-6 have increased the severity of knee OA. The relationship is positive, and the strength of the relationship is moderate at CTX-II (r = 0.546, CI 0.37–0.68), and IL-6 (r = 0.465 CI 0.27–0.62). For COMP, the relationship is positive and weak (r = 0.360, CI 0.15–0.54). The grade I-II knee osteoarthritis severity threshold is IL-6 of 2.14 pg/mL, COMP of 696.5 ng/mL, and CTX-II of 260.5 ng/mL.
ConclusionAutomatic assessment can detect initial knee OA (KL grades I and II). Biomarkers (IL-6, COMP, and CTX-II) can potentially detect knee OA in early stages.