QUANTITATIVE DIAGNOSIS OF OSTEOPOROSIS USING LUMBAR SPINE SIGNAL INTENSITY IN MAGNETIC RESONANCE IMAGING AMONG JORDANIAN POPULATION

Abstract

Background: Osteoporosis is a global health problem characterized by low bone mineral density (BMD), which may lead to greater bone fragility and fracture risk, especially in postmenopausal females. Magnetic resonance imaging (MRI) can be utilized, in addition to dual-energy X-ray absorptiometry (DXA). The purpose of our study was to determine the MRI-based score (M-score) corresponding to the T-score used in DXA and to assess the correlation between lumbar spine signal-to-noise ratio (SNR) measured by MRI and T-score among Jordanian postmenopausal women. Methods: Following approval by the Ethics Committee, we randomly selected 173 postmenopausal female patients who had DXA and lumbar-spine MRI, as well as a reference group of 34 premenopausal healthy adult patients (23-36 years) who had lumbar-spine MRI. In L1-L4, we calculated the SNR. Based on the T-score model, we proposed the M-score and the SNR. Receiver operator characteristic (ROC) analysis was utilized to identify the diagnostic performance of the M-score and SNR for osteoporosis detection with DXA. Results: Postmenopausal patients in the control group had significantly lower DXA scores overall and higher SNRL1 L4 and M-score than controls (p<0.0001). For differentiating between osteoporotic and non-osteoporotic females, the SNRL1-L4 diagnostic threshold was 184.2 with a sensitivity of 80% and specificity of 92%. The M-score diagnostic threshold was 3.5 with a sensitivity of 93.3% and specificity of 83.5%. A negative correlation between M-score and T score was detected. An optimal cut-off threshold of 5.95 was found for the M-score with a sensitivity of 80%, a specificity of 92%, and an accuracy of 95% for detecting osteoporotic patients from non-osteoporotic individuals. Conclusion: The lumbar spine MRI-based score (M-score) quantitative technique for differentiating osteoporotic from non-osteoporotic females. There was a strong correlation between osteoporotic indices (T-score and BMD) and M-score and SNRL1-L4.