OBJECTIVE: A prospective study to compare the MR and CT images of patients with suspected sacroiliitis and to establish the optimal MR sequences to demonstrate the changes of sacroiliitis was conducted. MATERIALS AND METHODS: Thirty-nine patients and nine controls were imaged in the axial plane, with SE T1-, T2-weighted fast spin echo (T2), T1 with fat suppression (T1WFS), and fast short tau inversion recovery (fast STIR) sequences on a 1.5 T system. The sacroiliac joints of all patients were imaged with CT. The images were evaluated by two independent radiologists. Following the blinded reading, direct comparison of T1 and T1WFS, T2, and fast STIR of the CT positive group was made to determine the optimal MR sequences. RESULTS: The sensitivity and specificity of MR images for the detection of cortical erosions and subchondral sclerosis when compared to CT images were 100 and 94.3%, respectively; interobserver variation was low (k = 0.80). T1WFS and fast STIR images were superior to T1 and T2 images, respectively, in demonstrating the changes of sacroiliitis. CONCLUSION: MRI (T1WFS and fast STIR) can replace CT in cases with a strong clinical suspicion of sacroiliitis and equivocal or normal plain radiographs.
OBJECTIVE: A prospective study to compare the MR and CT images of patients with suspected sacroiliitis and to establish the optimal MR sequences to demonstrate the changes of sacroiliitis was conducted. MATERIALS AND METHODS: Thirty-nine patients and nine controls were imaged in the axial plane, with SE T1-, T2-weighted fast spin echo (T2), T1 with fat suppression (T1WFS), and fast short tau inversion recovery (fast STIR) sequences on a 1.5 T system. The sacroiliac joints of all patients were imaged with CT. The images were evaluated by two independent radiologists. Following the blinded reading, direct comparison of T1 and T1WFS, T2, and fast STIR of the CT positive group was made to determine the optimal MR sequences. RESULTS: The sensitivity and specificity of MR images for the detection of cortical erosions and subchondral sclerosis when compared to CT images were 100 and 94.3%, respectively; interobserver variation was low (k = 0.80). T1WFS and fast STIR images were superior to T1 and T2 images, respectively, in demonstrating the changes of sacroiliitis. CONCLUSION: MRI (T1WFS and fast STIR) can replace CT in cases with a strong clinical suspicion of sacroiliitis and equivocal or normal plain radiographs.