To Örebro University

This work aims to evaluate whether radiological signs of stone impaction (RSSI) measured on non-contrast computed tomography (CT) can predict shock wave lithotripsy (SWL) outcomes for upper ureteral stones and to assess whether integrating these markers into an existing prediction model (the Niwa nomogram) improves predictive performance. We retrospectively analysed 256 patients treated with SWL for upper ureteral stones between 2012 and 2019. Standard stone parameters and RSSI, including ureteral wall thickness (UWT), ureteral diameters and CT attenuations above and below the stone, were assessed. Multivariable logistic regression, receiver operating characteristic (ROC) analysis, net reclassification improvement (NRI) and decision curve analysis (DCA) were used to evaluate predictive performance. The Niwa nomogram was enhanced by incorporating significant RSSI parameters and was internally validated using k-fold cross-validation. Maximum ureteral attenuation below the stone (UABSmax), ureter diameter above the stone (UDAS) and renal pelvis diameter (RPD) were found to be associated with SWL outcome. UABSmax had the highest individual predictive value (area under the curve (AUC) 0.66), while UWT showed no significant association or predictive value. Incorporating UABSmax and RPD into the Niwa nomogram (Niwa+) marginally increased AUC (0.72 vs. 0.71) but did not lead to significant improvements in NRI or DCA. In conclusion, certain RSSI- particularly UABSmax and RPD- were associated with SWL outcome but provided limited value when added to an already validated nomogram.

Stone size and location are key factors in predicting spontaneous stone passage (SSP), but little attention has been paid to the influence of radiological signs of stone impaction (RSSI). This research aims to determine whether RSSI, alongside stone size, can predict SSP and to evaluate the consistency of ureteral wall thickness (UWT) measurements among observers. In this retrospective study, 160 patients with a single upper or middle ureteral stone on acute non-enhanced computed tomography (NCCT) were analysed. Patient data were collected from medical records. Measurements of RSSI, including UWT, ureteral diameters, and average attenuation above and below the stone, were taken on NCCT by four independent readers blind to the outcomes. The cohort consisted of 70% males with an average age of 51 +/- 15. SSP occurred in 61% of patients over 20 weeks. The median stone length was 5.7 mm (IQR: 4.5-7.3) and was significantly shorter in patients who passed their stones at short- (4.6 vs. 7.1, p < 0.001) and long-term (4.8 vs. 7.1, p < 0.001) follow-up. For stone length, the area under the receiver operating characteristic curve (AUC) for predicting SSP was 0.90 (CI 0.84-0.96) and only increased to 0.91 (CI 0.85-0.95) when adding ureteral diameters and UWT. Ureteral attenuation did not predict SSP (AUC < 0.5). Interobserver variability for UWT was moderate, with +/- 2.0 mm multi-reader limits of agreement (LOA). The results suggest that RSSI do not enhance the predictive value of stone size for SSP. UWT measurements exhibit moderate reliability with significant interobserver variability.

Background: Viral infections, particularly Epstein-Barr virus (EBV), have been linked with risk of multiple sclerosis (MS). Given the evidence that SARS- CoV-2 infection can have consequences for the central nervous system (CNS) and autoimmune disorders, it might increase risk of MS and other demyelinating diseases of the CNS.

Objectives: We aimed to assess whether SARS- CoV-2 infection is associated with subsequent diagnoses of non-MS demyelinating CNS diseases, MS, and infectious mononucleosis (IM) due to EBV—an important MS risk factor.

Methods: All residents of Sweden aged 3–100 years were followed between 1st January 2020 and 30th November 2022, excluding those with demyelinating disease prior to 2020, resulting in 9,981,915 individuals. Exposure was classified as SARS- Cov-2 uninfected or infected, the latter divided by severity, and mod-elled as a time-varying covariate (uninfected, infection without hospital admission and infected with hospital admission). Cox regression assessed the risk of three separate outcomes: hospital-diagnosed non-MS demyelinating diseases; MS; and IM due t oEBV, adjusting for sex, year of birth (age), Charlson comorbidity index, healthcare region and country of birth.

Results: Hospital admission for COVID-19 was associated with raised risk of subsequent non-MS demyelinating disease. Rates per 100 000 person years (and 95% confidence intervals [CI]) were 3.8 (3.6– 4.1) among those without a COVID-19 diagnosis and 9.0 (5.1–15.9) among those admitted to hospital for COVID-19, with an adjusted hazard ratio (aHR) and 95% CI of 2.31 (1.30– 4.10). Equivalent associations with MS were rates of 9.5 (9.1–9.9) and 21.0 (14.5–30.5) per 100,000, and an aHR of 2.48 (1.70–3.61). For subsequent IM due to EBV, hospital admission for COVID-19 was associated with a rate of 10.5 (6.2–17.8) per 100,000 compared with 4.7 (4.4–5.0) for those without COVID-19, and an aHR of 5.63 (3.29–9.66).

Conclusions: There was increased risk of CNS demyelinating diseases among people admitted to hospital for COVID-19. COVID-19 was also associated with a raised risk of IM due to EBV, an established risk factor for MS. It is possible that at least a proportion of these associations is due to surveillance or referral bias (due to a previous hospital admission for infection), so future research should continue to follow the population that had COVID-19 for development of MS and other demyelinating diseases, which can have long asymptomatic and prodromal phases.

Demyelinating diseases including multiple sclerosis are associated with prior infectious exposures, so we assessed whether SARS-CoV-2 infection is associated with subsequent diagnoses of non-multiple sclerosis demyelinating diseases and multiple sclerosis. All residents of Sweden aged 3-100 years were followed between 1 January 2020 and 30 November 2022, excluding those with demyelinating disease prior to 2020, comprising 9 959 818 individuals divided into uninfected and those who were infected were categorized into those with and without hospital admission for the infection as a marker of infection severity. Cox regression assessed the risk of two separate outcomes: hospital diagnosed non-multiple sclerosis demyelinating diseases of the CNS and multiple sclerosis. The exposures were modelled as time-varying covariates (uninfected, infection without hospital admission and infected with hospital admission). Hospital admission for COVID-19 was associated with raised risk of subsequent non-multiple sclerosis demyelinating disease, but only 12 individuals had this outcome among the exposed, and of those, 7 has an unspecified demyelinating disease diagnosis. Rates per 100 000 person-years (and 95% confidence intervals) were 3.8 (3.6-4.1) among those without a COVID-19 diagnosis and 9.0 (5.1-15.9) among those admitted to hospital for COVID-19, with an adjusted hazard ratio and (and 95% confidence interval) of 2.35 (1.32-4.18, P = 0.004). Equivalent associations with multiple sclerosis (28 individuals had this outcome among the exposed) were rates of 9.5 (9.1-9.9) and 21.0 (14.5-30.5) and an adjusted hazard ratio of 2.48 (1.70-3.61, P < 0.001). Only a small number of non-multiple sclerosis demyelinating disease diagnoses were associated with hospital admission for COVID-19, and while the number with multiple sclerosis was somewhat higher, longer duration of follow-up will assist in identifying whether the associations are causal or due to shared susceptibility or surveillance bias, as these diseases can have long asymptomatic and prodromal phases.

We aimed to externally validate five previously published predictive models (Ng score, Triple D score, S3HoCKwave score, Kim nomogram, Niwa nomogram) for shock wave lithotripsy (SWL) single-session outcomes in patients with a solitary stone in the upper ureter. The validation cohort included patients treated with SWL from September 2011 to December 2019 at our institution. Patient-related variables were retrospectively collected from the hospital records. Stone-related data including all measurements were retrieved from computed tomography prior to SWL. We estimated discrimination using area under the curve (AUC), calibration, and clinical net benefit based on decision curve analysis (DCA). A total of 384 patients with proximal ureter stones treated with SWL were included in the analysis. Median age was 55.5 years, and 282 (73%) of the sample were men. Median stone length was 8.0 mm. All models significantly predicted the SWL outcomes after one session. S3HoCKwave score, Niwa, and Kim nomograms had the highest accuracy in predicting outcomes, with AUC 0.716, 0.714 and 0.701, respectively. These three models outperformed both the Ng (AUC: 0.670) and Triple D (AUC: 0.667) scoring systems, approaching statistical significance (P = 0.05). Of all the models, the Niwa nomogram showed the strongest calibration and highest net benefit in DCA. To conclude, the models showed small differences in predictive power. The Niwa nomogram, however, demonstrated acceptable discrimination, the most accurate calibration, and the highest net benefit whilst having relatively simple design. Therefore, it could be useful for counselling patients with a solitary stone in the upper ureter.

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.

Computed tomography (CT) is the mainstay imaging method for urinary stones.

The aim of this thesis was to optimize the information obtained from the initial CT scan to allow a well-founded diagnosis and prognosis, and to guide the clinician as early and as far as possible in the further treatment of urinary stone disease.

We examined CT scan parameters with regards to their importance for prediction of spontaneous ureteral stone passage, the impact of interreader variability of stone size estimates on this prediction, and the predictive accuracy of a semi-automated, three-dimensional (3D) segmentation algorithm. We also developed and tested the ability of a machine learning algorithm to classify pelvic calcifications into ureteral stones and phleboliths.

Using single-energy CT, three quantitative methods for classification of stone composition into uric acid and non-uric acid stones in vivo were prospectively validated, using dual-energy CT as reference.

Our results show that spontaneous ureteral stone passage can be predicted with high accuracy, with knowledge of stone size and position. The interreader variability in the size estimation has a large impact on the predicted outcome, but can be eliminated through a 3D segmentation algorithm. Which size estimate we use is of minor importance, but it is important that we use the chosen estimate consistently. A machine learning algorithm can differentiate distal ureteral stones from phleboliths, but more than local features are needed to reach optimal discrimination.

A single-energy CT method can distinguish uric acid from non-uric acid stones in vivo with accuracy comparable to dual-energy CT.

In conclusion, single-energy CT not only detects a urinary stone, but can also provide us with a prediction regarding spontaneous stone passage and a classification of stone type into uric acid and non-uric acid.

OBJECTIVES: To prospectively validate three quantitative single-energy CT (SE-CT) methods for classifying uric acid (UA) and non-uric acid (non-UA) stones.

METHODS: Between September 2018 and September 2019, 116 study participants were prospectively included in the study if they had at least one 3-20-mm urinary stone on an initial urinary tract SE-CT scan. An additional dual-energy CT (DE-CT) scan was performed, limited to the stone of interest. Additionally, to include a sufficient number of UA stones, eight participants with confirmed UA stone on DE-CT were retrospectively included. The SE-CT stone features used in the prediction models were (1) maximum attenuation (maxHU) and (2) the peak point Laplacian (ppLapl) calculated at the position in the stone with maxHU. Two prediction models were previously published methods (ppLapl-maxHU and maxHU) and the third was derived from the previous results based on the k-nearest neighbors (kNN) algorithm (kNN-ppLapl-maxHU). The three methods were evaluated on this new independent stone dataset. The reference standard was the CT vendor's DE-CT application for kidney stones.

RESULTS: Altogether 124 participants (59 ± 14 years, 91 men) with 106 non-UA and 37 UA stones were evaluated. For classification of UA and non-UA stones, the sensitivity, specificity, and accuracy were 100% (37/37), 97% (103/106), and 98% (140/143), respectively, for kNN-ppLapl-maxHU; 95% (35/37), 98% (104/106), and 97% (139/143) for ppLapl-maxHU; and 92% (34/37), 94% (100/106), and 94% (134/143) for maxHU.

CONCLUSION: A quantitative SE-CT method (kNN-ppLapl-maxHU) can classify UA stones with accuracy comparable to DE-CT.

KEY POINTS:

• Single-energy CT is the first-line diagnostic tool for suspected renal colic.

• A single-energy CT method based on the internal urinary stone attenuation distribution can classify urinary stones into uric acid and non-uric acid stones with high accuracy.

• This immensely increases the availability of in vivo stone analysis.

Computed tomography (CT) is the method of choice for diagnosing ureteral stones - kidney stones that obstruct the ureter. The purpose of this study is to develop a computer aided detection (CAD) algorithm for identifying a ureteral stone in thin slice CT volumes. The challenge in CAD for urinary stones lies in the similarity in shape and intensity of stones with non-stone structures and how to efficiently deal with large high-resolution CT volumes. We address these challenges by using a Convolutional Neural Network (CNN) that works directly on the high resolution CT volumes. The method is evaluated on a large data base of 465 clinically acquired high-resolution CT volumes of the urinary tract with labeling of ureteral stones performed by a radiologist. The best model using 2.5D input data and anatomical information achieved a sensitivity of 100% and an average of 2.68 false-positives per patient on a test set of 88 scans.

Purpose: To develop a deep learning method for assisting radiologists in the discrimination between distal ureteral stones and pelvic phleboliths in thin slice CT images, and to evaluate whether this differentiation is possible using only local features.

Methods and materials: A limited field-of-view image data bank was retrospectively created, consisting of 5x5x5 cm selections from 1 mm thick unenhanced CT images centered around 218 pelvis phleboliths and 267 distal ureteral stones in 336 patients. 50 stones and 50 phleboliths formed a validation cohort and the remainder a training cohort. Ground truth was established by a radiologist using the complete CT examination during inclusion.The limited field-of-view CT stacks were independently reviewed and classified as containing a distal ureteral stone or a phlebolith by seven radiologists. Each cropped stack consisted of 50 slices (5x5 cm field-of-view) and was displayed in a standard PACS reading environment. A convolutional neural network using three perpendicular images (2.5D-CNN) from the limited field-of-view CT stacks was trained for classification.

Results: The 2.5D-CNN obtained 89% accuracy (95% confidence interval 81%-94%) for the classification in the unseen validation cohort while the accuracy of radiologists reviewing the same cohort was 86% (range 76%-91%). There was no statistically significant difference between 2.5D-CNN and radiologists.

Conclusion: The 2.5D-CNN achieved radiologist level classification accuracy between distal ureteral stones and pelvic phleboliths when only using the local features. The mean accuracy of 86% for radiologists using limited field-of-view indicates that distant anatomical information that helps identifying the ureter’s course is needed.