Comparison of real-time ultrasonography and magnetic resonance imaging in the assessment of urinary bladder volume.

Abstract

To compare estimates of bladder volume obtained by conventional real-time ultrasonography with those obtained from magnetic resonance imaging (MRI) by the Cavalieri method of unbiased stereology. The study comprised nine subjects (four men and five women, mean age 23 years, range 18-34) with no history of bladder disease. Before micturition, each volunteer underwent ultrasonography, immediately followed by MRI. The volunteers then voided the true voided volume of urine was measured and the imaging protocols were repeated in the same order after micturition. The bladder volume was estimated from ultrasonography using the formula: volume = 0.7 (L x TS x AP), (where L is the maximum supero-inferior diameter. AP the maximum anteroposterior diameter and TS the maximum transverse diameter) and from MRI using the Cavalieri method. For each imaging modality, the volume of urine voided was estimated as the difference in the volume estimate before and after micturition. The mean percentage coefficient of variation for the estimates of bladder volume by ultrasonography was 2.17 before and 4.43 after micturition. There was no significant difference in the replicate estimates of each bladder diameter by ultrasonography before and after micturition (P = 0.98). The MRI method consistently underestimated the voided volume: the mean discrepancy between the estimated voided volume and the true voided volume was 7.7 ml, and -67.7 ml for the ultrasonographic and MRI estimates, respectively, which are significantly different (P = 0.02) when assessed using a multifactor ANOVA. Further analysis using multiple-range tests showed a significant difference between the voided volume estimated by MRI and the corresponding true voided volume. There was no difference between the voided volume estimated by ultrasonography and the corresponding true volume. Ultrasonographic estimates of voided volume were more reliable than the those obtained using the MRI method. This is possibly due to a delay between micturition and the acquisition of MR images after micturition, which allowed the bladder to partly refill with urine. The empirical approach using measurements from ultrasonograms provides a fast and reliable technique: ultrasonography remains the recommended imaging modality for estimating bladder volume.