Comparison of the diagnostic accuracy of diffusion-weighted and dynamic contrast-enhanced MRI with 18F-FDG PET/CT to differentiate osteomyelitis from Charcot neuro-osteoarthropathy in diabetic foot

Abstract

Purpose: To compare the diagnostic accuracy of diffusion-weighted imaging (DWI) and dynamic contrast-enhanced-magnetic resonance imaging (DCE-MRI) involving two region of interest (ROI) sizes with 18-fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG PET/CT) to differentiate diabetic foot osteomyelitis (DFO) from Charcot neuro-osteoarthropathy (CN). Method: Thirty-one diabetic patients were included in this prospective study. Two readers independently evaluated DWI (apparent diffusion coefficient [ADC] and high-b-value signal pathological-to-normal bone ratio [DWIr]) and DCE-MRI parameters (Ktrans, Kep, Ve, internal area under the gadolinium curve at 60 s [iAUC60] and time intensity curve [TIC]) using two different ROI sizes, and 18F-FDG PET/CT parameters (visual assessment, SUVmax, delayed SUVmax, and percentage changes between SUVmax and delayed SUVmax). Techniques were compared by univariate analysis using the area under the receiver operating characteristic curve [AUC]. Reliability was analyzed with Kappa and Intraclass correlation [ICC]. Results: DWIr, Ktrans and iAUC60 showed better diagnostic accuracy (AUC = 0.814−0.830) and reliability (ICC > 0.9) for large than for small ROIs (AUC = 0.736−0.750; ICC = 0.6 in Ktrans, 0.8 in DWIr and iAUC60). TIC showed moderate diagnostic performance (AUC = 0.739−0.761) and reliability (κ 0.7). Visual assessment of 18F-FDG PET/CT demonstrated a significantly higher accuracy (AUC = 0.924) than MRI parameters. Semi-quantitative 18F-FDG PET/CT parameters did not provide significant improvement over visual analysis (AUC = 0.848−0.903). Conclusion: DWIr, Ktrans and iAUC60 allowed reliable differentiation of DFO and CN, particularly for large ROIs. Visual assessment of 18F-FDG PET/CT was the most accurate technique for differentiation.