Login to Access Video or Poster Abstract: MP01-08
Sources of Funding: Mayo Clinic Urolithiasis O'Brien Grant

Introduction

Computed Tomography (CT) is a clinically established modality to evaluate suspected urinary stones. The maximum stone dimension in the axial reconstruction and stone location are often used to estimate the probability of spontaneous stone passage and potential likelihood of surgical intervention. However, the measured axial dimension of urinary stones can vary considerably owing to irregular shape, obliquity to the imaging plane, non-isotropic imaging voxels, interobserver variability, and volume averaging. This limits the reproducibility of axial stone measurements and the accuracy of predictions based upon maximum axial stone dimension. The present study compared the standard measures of stone size from axial images to those obtained using dedicated stone analysis software, which determined maximal stone dimensions in all planes.

Methods

Non-contrast-enhanced abdominal CT scans from 211 consecutive emergency department patients performed to evaluate flank were retrospectively evaluated. Radiological reports were reviewed for a diagnosis of urolithiasis, the maximum axial stone dimension, and stone location. Corresponding 1 mm thick images were analyzed using dedicated stone analysis software to compute the maximum linear dimension in any direction and stone volume. Descriptive outcomes are reported here (mean (SD)), comparing traditional maximum axial dimension and stone volume (assuming a spherical stone) to measurements made using dedicated software that performed 3D stone segmentation.

Results

A total of 228 stones were identified in 143 of the 211 patients. The mean maximum dimension in any direction computed by the software algorithm was 5.0 (3.2) mm, which was significantly higher than the mean maximum dimension of 3.9 (2.9) mm contained in the radiographic reports (p=0.0002). The actual stone volume computed by the algorithm based upon the true stone dimensions and shape was 52.8 (141.5) mm3, while the stone volume calculated assuming a spherical shape was 31.06 (102.16) mm3 (p=0.0628).

Conclusions

Using dedicated stone analysis software, maximal stone dimension in any plane and stone volume were significantly larger than traditional measurements made in the axial plane and the associated volume. Semi-automated 3D measurements of stone size hence may be more accurate and reproducible. Further studies are needed to determine if automated 3D stone size metrics offer improved and more reliable prediction of spontaneous stone passage.

Funding

Mayo Clinic Urolithiasis O'Brien Grant