To Örebro University

OBJECTIVES: To determine how to most accurately predict the chance of spontaneous passage of a ureteral stone using information in the diagnostic non-enhanced computed tomography (NECT) and to create predictive models with smaller stone size intervals than previously possible.

METHODS: Retrospectively 392 consecutive patients with ureteric stone on NECT were included. Three radiologists independently measured the stone size. Stone location, side, hydronephrosis, CRP, medical expulsion therapy (MET) and all follow-up radiology until stone expulsion or 26 weeks were recorded. Logistic regressions were performed with spontaneous stone passage in 4 weeks and 20 weeks as the dependent variable.

RESULTS: The spontaneous passage rate in 20 weeks was 312 out of 392 stones, 98% in 0-2 mm, 98% in 3 mm, 81% in 4 mm, 65% in 5 mm, 33% in 6 mm and 9% in ≥6.5 mm wide stones. The stone size and location predicted spontaneous ureteric stone passage. The side and the grade of hydronephrosis only predicted stone passage in specific subgroups.

CONCLUSION: Spontaneous passage of a ureteral stone can be predicted with high accuracy with the information available in the NECT. We present a prediction method based on stone size and location.

KEY POINTS: • Non-enhanced computed tomography can predict the outcome of ureteral stones. • Stone size and location are the most important predictors of spontaneous passage. • Prediction models based on stone width or length and stone location are introduced. • The observed passage rates for stone size in mm-intervals are reported. • Clinicians can make better decisions about treatment.

OBJECTIVES: To compare the ability of different size estimates to predict spontaneous passage of ureteral stones using a 3D-segmentation and to investigate the impact of manual measurement variability on the prediction of stone passage.

METHODS: We retrospectively included 391 consecutive patients with ureteral stones on non-contrast-enhanced CT (NECT). Three-dimensional segmentation size estimates were compared to the mean of three radiologists' measurements. Receiver-operating characteristic (ROC) analysis was performed for the prediction of spontaneous passage for each estimate. The difference in predicted passage probability between the manual estimates in upper and lower stones was compared.

RESULTS: The area under the ROC curve (AUC) for the measurements ranged from 0.88 to 0.90. Between the automated 3D algorithm and the manual measurements the 95% limits of agreement were 0.2 ± 1.4 mm for the width. The manual bone window measurements resulted in a > 20 percentage point (ppt) difference between the readers in the predicted passage probability in 44% of the upper and 6% of the lower ureteral stones.

CONCLUSIONS: All automated 3D algorithm size estimates independently predicted the spontaneous stone passage with similar high accuracy as the mean of three readers' manual linear measurements. Manual size estimation of upper stones showed large inter-reader variations for spontaneous passage prediction.

KEY POINTS:• An automated 3D technique predicts spontaneous stone passage with high accuracy.• Linear, areal and volumetric measurements performed similarly in predicting stone passage.• Reader variability has a large impact on the predicted prognosis for stone passage.

The objectives were to develop and validate a Convolutional Neural Network (CNN) using local features for differentiating distal ureteral stones from pelvic phleboliths, compare the CNN method with a semi-quantitative method and with radiologists' assessments and to evaluate whether the assessment of a calcification and its local surroundings is sufficient for discriminating ureteral stones from pelvic phleboliths in non-contrast-enhanced CT (NECT). We retrospectively included 341 consecutive patients with acute renal colic and a ureteral stone on NECT showing either a distal ureteral stone, a phlebolith or both. A 2.5-dimensional CNN (2.5D-CNN) model was used, where perpendicular axial, coronal and sagittal images through each calcification were used as input data for the CNN. The CNN was trained on 384 calcifications, and evaluated on an unseen dataset of 50 stones and 50 phleboliths. The CNN was compared to the assessment by seven radiologists who reviewed a local 5 × 5 × 5 cm image stack surrounding each calcification, and to a semi-quantitative method using cut-off values based on the attenuation and volume of the calcifications. The CNN differentiated stones and phleboliths with a sensitivity, specificity and accuracy of 94%, 90% and 92% and an AUC of 0.95. This was similar to a majority vote accuracy of 93% and significantly higher (p = 0.03) than the mean radiologist accuracy of 86%. The semi-quantitative method accuracy was 49%. In conclusion, the CNN differentiated ureteral stones from phleboliths with higher accuracy than the mean of seven radiologists' assessments using local features. However, more than local features are needed to reach optimal discrimination.