Objectives: The aim of the study was to compare the inter-observer variability and the accuracy of magnetic resonance urography (MRU) using a thin sectional balanced-turbo field echo (B-TFE) sequence for detecting ureteral calculi and to determine the effect of additional factors (size, density and location of the calculus) on the sensitivity and specificity of the MRU.
Materials & methods: MRU and CT images were evaluated independently by two radiologists according to presence, density and localization of calculi. The degrees of inter-rater agreement for categorical items were evaluated by the Kappa coefficient.
Results: According to the 1st and 2nd observers, the sensitivity of MRU was 65.9 %, 71.8 % and the specificity of MRU was 95.9 %, 100 %, respectively. Inter-observer agreement was 84.6 % for stone detection. The larger size had a better effect on detectability (p < 0.05). Also, the higher density had a better impact on detectability (p < 0.05).
Conclusion: Our study has shown that B-TFE MRU was useful to detect ureteral calculi. However, B-TFE MRU has low sensitivity and high specificity in comparison with CT images. MRU is a reasonable alternative imaging technique for follow-up periods of selective groups like patients with large urinary stones, children or pregnant patients when ionizing radiation is undesirable.
Key points: • According to 1st and 2nd observers, sensitivity of MRU was 65.9 %, 71.8 %, respectively. • According to 1st and 2nd observers, MRU specificity was 95.9 %, 100 %, respectively. • Interobserver agreement was found to be over 84 % for stone detection. • B-TFE sequence provides calculus follow-up without radiation. • Larger calculi and more dense calculi individually have the better effect on detectability.
Keywords: Balanced-turbo field echo; Diagnosis; Flank pain; Magnetic resonance urography; Ureteral calculus.