#0694

The efficacy of the attenuation value of bladder urine on non-contrast CT for predicting urine culture positivity

G. Kaneko¹, S. Ikuma¹, R. Matsuoka¹, J. Akatsuka², Y. Ogata¹, M. Sato¹, Y. Kondo²

¹Nippon Medical School Tamanagayama Hospital, Department of Urology, Tokyo, Japan
²Nippon Medical School Hospital, Department of Urology, Tokyo, Japan

Introduction:

To evaluate the predictive value of bladder urine attenuation (HU; Hounsfield units) on non-contrast CT for detecting urine culture positivity in patients with urinary tract obstruction secondary to ureteral stones.

Material and methods:

This study included patients who were hospitalized for a urinary tract infection associated with hydronephrosis secondary to ureteral stones confirmed by non-contrast CT between December 2020 and September 2024 at our hospital. Patients who did not undergo urine culture testing were excluded. We manually segmented the renal pelvis and bladder using 3D Slicer (version 4.8.1) to create volumes of interest (VOI). We measured the volume, attenuation value (HU), and coefficient of variation (CV) from each VOI. We then analyzed the correlation between each of these factor and urine culture positivity. Receiver operating characteristic (ROC) curve analysis was then performed to determine the optimal cutoff value for predicting urine culture positivity.

Results:

This study included 71 patients (39 men, 32 women; median age, 70 years), of whom 28 (39.4%) had a positive urine culture. The attenuation value of bladder urine was significantly lower in the urine culture positive group compared to the negative group (6.53 vs. 12.80, OR: 0.891, 95% CI: 0.82–0.96, P = 0.004). There were no significant differences in urine volume or CV between the two groups. ROC curve analysis showed an AUC of 0.72 for bladder urine HU, and the optimal cutoff value was 10.2 HU.