Value of MRI signal intensity in evaluation of allergic fungal rhinosinusitis compared with CT Hounsfield units: Retrospective study

Abstract

“Allergic fungal sinusitis (AFS)” is typically diagnosed using radiologic images like computed tomography (CT) scans and magnetic resonance imaging (MRI), with the “Hounsfield unit (HU)” in CT scans and T2-weighted images (T2WI) in MRI serving as reliable objective parameters. However, diagnosing AFS might be difficult because of possible signal changes and densities caused by variations in the secretion concentration in the sinus. Few studies have compared the diagnostic performance of MRI and CT scans. This study aimed to investigate the value of MRI signal intensity in evaluating AFS compared with CT HUs. This retrospective study included 111 patients with pathologically confirmed AFS who underwent CT imaging followed by MRI evaluation at King Saud Medical City, Riyadh, Saudi Arabia, from January 2012 to December 2022. Radiographic densities of sinus opacities on CT scan, including the mean HU values, and MRI findings, including signal voids on T1-weighted images and T2WI, were gathered and analyzed. To determine the efficacy of these radiographic characteristics in predicting the disease and the best cutoff value, we employed receiver operator characteristic curves. The mean age was 31.9 ± 15.6 years, and most patients were 74 females (66.7%). The main symptom was nasal obstruction in 73 patients (65.8%). In comparison, between HU and signal void on T2WI, there was moderate predictive performance [area under the curve: 0.856, P = .001]. An ideal HU cutoff value of 69.50 HU was obtained with a sensitivity of 100% and a specificity of 44.7%. However, the receiver operator characteristic for T1-weighted images could not be plotted, as no signal was avoided to predict AFS and it was not statistically significant (area under the curve: 0.566; P = .287). The study found a CT HU of 69.5 can predict MRI T2WI signal values with a void signal, aiding in diagnostic workup and evaluation for AFS.