Assessment of urinary tract calculi with 64-MDCT: The axial versus coronal plane

ACR Appropriateness Criteria® Acute Onset Flank Pain-Suspicion of Stone Disease (Urolithiasis)

Noncontrast CT (NCCT) is the imaging study of choice for initial evaluation of patients with acute onset of flank pain and suspicion of stone disease without known prior stone disease. NCCT can reliably characterize the location and size of an offending ureteral calculus, identify complications, and diagnose alternative etiologies of abdominal pain. Although less sensitive in the detection of stones, ultrasound may have a role in evaluating for signs of obstruction. Radiography potentially has a role, although has been shown to be less sensitive than NCCT. For patients with known disease and recurrent symptoms of urolithiasis, NCCT remains the test of choice for evaluation. In pregnancy, given radiation concerns, ultrasound is recommended as the initial modality of choice with potential role for noncontrast MRI. In scenarios where stone disease suspected and initial NCCT is inconclusive, contrast-enhanced imaging, either with MRI or CT/CT urogram may be appropriate. The American College of Radiology Appropriateness Criteria are evidence-based guidelines for specific clinical conditions that are reviewed annually by a multidisciplinary expert panel. The guideline development and revision process support the systematic analysis of the medical literature from peer reviewed journals. Established methodology principles such as Grading of Recommendations Assessment, Development, and Evaluation or GRADE are adapted to evaluate the evidence. The RAND/UCLA Appropriateness Method User Manual provides the methodology to determine the appropriateness of imaging and treatment procedures for specific clinical scenarios. In those instances where peer reviewed literature is lacking or equivocal, experts may be the primary evidentiary source available to formulate a recommendation.

[Imaging for urolithiasis]

The substantial reduction of radiation exposure using (ultra-)low dose programs in native computed tomographic imaging has led to considerable changes in imaging diagnostics and treatment planning in urolithiasis in recent years. In addition, especially in Germany, ultrasound diagnostics is highly available in terms of equipment and with increasing expertise. This can largely replace the previous radiation-associated procedures in emergency and follow-up diagnostics, but also in intraoperative imaging, e.g., in percutaneous stone therapy (intraoperative fluoroscopy). This is reflected in the international guidelines, which recommend these two modalities as first-line diagnostics in all areas mentioned. Continuous technical development enables ever higher resolution imaging and thus improved diagnostics with high sensitivity and specificity. This also enables reliable imaging of particularly vulnerable patient groups, such as children or pregnant women. In addition, methods from the field of artificial intelligence (AI; machine learning, deep learning) are increasingly being used for automated stone detection and stone characterization including its composition. Furthermore, AI models can provide prognosis models as well as individually tailored treatment, follow-up, and prophyaxis. This will enable further personalization of diagnostics and therapy in the field of urolithiasis.

Urinary Stone Detection on CT Images Using Deep Convolutional Neural Networks: Evaluation of Model Performance and Generalization

PURPOSE

To investigate the diagnostic accuracy of cascading convolutional neural network (CNN) for urinary stone detection on unenhanced CT images and to evaluate the performance of pretrained models enriched with labeled CT images across different scanners.

MATERIALS AND METHODS

This HIPAA-compliant, institutional review board-approved, retrospective clinical study used unenhanced abdominopelvic CT scans from 535 adults suspected of having urolithiasis. The scans were obtained on two scanners (scanner 1 [hereafter S1] and scanner 2 [hereafter S2]). A radiologist reviewed clinical reports and labeled cases for determination of reference standard. Stones were present on 279 (S1, 131; S2, 148) and absent on 256 (S1, 158; S2, 98) scans. One hundred scans (50 from each scanner) were randomly reserved as the test dataset, and the rest were used for developing a cascade of two CNNs: The first CNN identified the extent of the urinary tract, and the second CNN detected presence of stone. Nine variations of models were developed through the combination of different training data sources (S1, S2, or both [hereafter SB]) with (ImageNet, GrayNet) and without (Random) pretrained CNNs. First, models were compared for generalizability at the section level. Second, models were assessed by using area under the receiver operating characteristic curve (AUC) and accuracy at the patient level with test dataset from both scanners (n = 100).

RESULTS

The GrayNet-pretrained model showed higher classifier exactness than did ImageNet-pretrained or Random-initialized models when tested by using data from the same or different scanners at section level. At the patient level, the AUC for stone detection was 0.92-0.95, depending on the model. Accuracy of GrayNet-SB (95%) was higher than that of ImageNet-SB (91%) and Random-SB (88%). For stones larger than 4 mm, all models showed similar performance (false-negative results: two of 34). For stones smaller than 4 mm, the number of false-negative results for GrayNet-SB, ImageNet-SB, and Random-SB were one of 16, three of 16, and five of 16, respectively. GrayNet-SB identified stones in all 22 test cases that had obstructive uropathy.

CONCLUSION

A cascading model of CNNs can detect urinary tract stones on unenhanced CT scans with a high accuracy (AUC, 0.954). Performance and generalization of CNNs across scanners can be enhanced by using transfer learning with datasets enriched with labeled medical images.© RSNA, 2019Supplemental material is available for this article. : An earlier incorrect version appeared online. This article was corrected on August 6, 2019.

A radiologist’s approach to CT KUB for the urologist

In this article we review the computed tomography imaging features of renal and ureteric calculi, the secondary signs of renal tract calculi and their possible complications. Most calculi appear as high-density foci due to their composition, whereas other signs that may be appreciated include the soft-tissue rim sign in the ureter, upstream renal tract dilation and peri-ureteric or peri-renal stranding. The most accurate means of measuring a calculus is in the coronal plane on bone windows. We also provide a systematic approach to interrogating computed tomography of the kidneys, ureter and bladder for the urologist, with examples of other conditions that may present as renal colic and are apparent on computed tomography.

Level of evidence: Not applicable as no original research is being proffered.

Characteristics of gout patients according to the laterality of nephrolithiasis: A cross-sectional study using helical computed tomography

OBJECTIVE

To clarify the clinical and laboratory characteristics of nephrolithiasis in gout by computed tomography (CT).

METHODS

In 350 gout patients, unenhanced CT was performed at the 1st visit to hospital. Calculus density spots exceeding 1 mm in diameter with a CT value >120 Hounsfield units in the kidneys were defined as kidney stones. The association between laterality and the number of stones was investigated in each stone carrier. The 350 patients were classified into three groups (bilateral, unilateral and non-stone carriers). Then serum urate (Sua), renal function, uric acid metabolism, and the prevalence of metabolic syndrome (Mets) were compared among these groups by the Tukey-Kramer test or Fisher's exact test.

RESULTS

Kidney stone(s) were detected in 108 (31%) of the 350 patients (bilateral in 58 and unilateral in 50). In 64 of the 108 patients (59%), there was no history of urolithiasis. Sua, serum creatinine and uric acid clearance were significantly higher (P = 0.001, P < 0.001, P = 0.043, respectively), while the estimated glomerular filtration rate was significantly lower (P = 0.039) in bilateral stone carriers than in non-stone carriers. No significant differences of uric acid metabolism or the prevalence of Mets were noted among the three groups.

CONCLUSIONS

Approximately one-third of gout patients had kidney stones and more than half of the patients with stones were bilateral and multiple stone carriers. Elevation of Sua might increase the stone burden in gout, leading to more severe renal dysfunction. An association between nephrolithiasis and Mets was not demonstrated in gout patients.

Ureteral wall thickness as a significant factor in predicting spontaneous passage of ureteral stones of ≤ 10 mm: a preliminary report

PURPOSE

To evaluate the clinical significance of ureteral wall thickness (UWT) for predicting spontaneous passage (SP) of uncomplicated ureteral stones of ≤ 10 mm.

METHODS

We retrospectively reviewed 418 patients with a diagnosis of uncomplicated ureteral stones who presented to Kori Hospital from 2011 to 2018. The maximum stone diameter and UWT at the stone site were measured from axial computed tomography images. Clinical predictors of 4-week SP were assessed using univariate and multivariate analyses. Receiver operating characteristic (ROC) curve analysis was applied to evaluate the accuracy of factors in predicting SP of ureteral stones.

RESULTS

Of the 418 patients, 202 (48.3%) spontaneously passed their stones within 4 weeks of their initial visit. Multivariate analysis showed that stone location, stone size, and UWT (odds ratio, 0.40; P < 0.001) were independent predictors of 4-week SP. ROC analysis showed that 2.71 mm was the optimal cut-off value for UWT, with a predictive accuracy of 0.83. Low UWT had a significantly higher 4-week SP rate than high UWT (76.4% vs. 14.7%, respectively; P < 0.001). In addition, high UWT was associated with a higher risk of stone-related complications within 4 weeks compared with low UWT (16.4% vs. 7.2%, respectively; P = 0.0044).

CONCLUSIONS

UWT can serve as a potential predictive factor for 4-week SP and may help physicians to select patients who require immediate interventions among those with ≤ 10-mm ureteral stones.

Assessment of factors affecting the spontaneous passage of lower ureteric calculus on the basis of lower ureteric calculus diameter, density, and plasma C- reactive protein level

Introduction:

The study aimed to evaluate the factors which affect the spontaneous passage of lower ureteric calculus on the basis of noncontrast computed tomography kidneys, ureters, and bladder (NCCT KUB) stone diameter, stone density, and plasma C-reactive protein (CRP) level.

Materials and Methods:

We conducted a prospective study of 200 patients with lower ureteric calculus 5–10 mm in size, from October 2015 to December 2016. All patients underwent NCCT KUB region with a 5 mm axial and reformatted coronal section. Edema just above the calculus and rim sign at the level of calculus and density of calculus is evaluated. Only scan with isolated, unilateral, solitary ureteric calculus was included in the final analysis and monitored up to 4 weeks, and plasma CRP is estimated in all patients to determine the clinical outcome.

Results:

A total of 200 patients (145 males, 55 females; mean age ± standard deviation, 34.73 ± 10.29) were included in the study. Lower ureteric calculus between 5–7 mm passed in 70% and 7–10 mm passed in 40%. There was 18% underestimation of maximum stone diameter in axial plane as compared to coronal plane. For spontaneous passage of calculus, craniocaudal (CC) diameter is more reliable then axial in NCCT. Rim sign and edema is absent in 64% of those passed spontaneous calculus. CRP level more than 2.45 mg/dl has low spontaneous expulsion rate. The stone with different HU passes through the ureter with same rate.

Conclusion:

Plasma CRP level and CC diameter and absence of rim sign on NCCT KUB are more reliable factors then density for spontaneous passage of ureteric calculus.

Factors affecting the urologist's decision to administer ureteral stone therapy: a retrospective cohort study

Objective

We aimed to evaluate the factors influencing treatment option selection among urologists for patients with ureteral stones, according to the stone diameter and location.

Methods

We retrospectively reviewed the records of 360 consecutive patients who, between January 2009 and June 2014, presented to the emergency department with renal colic and were eventually diagnosed with urinary stones via computed tomography. The maximal horizontal and longitudinal diameter and location of the stones were investigated. We compared parameters between patients who received urological intervention (group 1) and those who received medical treatment (group 2).

Results

Among the 360 patients, 179 (49.7%) had stones in the upper ureter and 181 (50.3%) had stones in the lower ureter. Urologic intervention was frequently performed in cases of upper ureteral stones (P<0.001). In groups 1 and 2, the stone horizontal diameters were 5.5 mm (4.8 to 6.8 mm) and 4.0 mm (3.0 to 4.6 mm), stone longitudinal diameters were 7.5 mm (6.0 to 9.5 mm) and 4.4 mm (3.0 to 5.5 mm), and ureter diameters were 6.4 mm (5.0 to 8.0 mm) and 4.7 mm (4.0 to 5.3 mm), respectively (P<0.001). The cut-off values for the horizontal and longitudinal stone diameters in the upper ureter were 4.45 and 6.25 mm, respectively (sensitivity 81.3%, specificity 91.4%); those of the lower ureter were 4.75 and 5.25 mm, respectively (sensitivity 79.4%, specificity 79.4%).

Conclusion

The probability of a urologic intervention was higher for patients with upper ureteral stones and those with stone diameters exceeding 5 mm horizontally and 6 mm longitudinally.

Variation in Radiologic and Urologic Computed Tomography Interpretation of Urinary Tract Stone Burden: Results From the Registry for Stones of the Kidney and Ureter

OBJECTIVE

To compare the measured stone burden recorded between urologists and radiologists, and examine how these differences could potentially impact stone management. As current urologic stone surgery guideline recommendations are based on stone size, accurate stone measurements are crucial to direct appropriate treatment. This study investigated the discrepant interpretation that often exists between urologic surgeons and radiologists' estimation of patient urinary stone burden.

MATERIALS AND METHODS

From November 2015 through August 2016, new patients prospectively enrolled into the Registry for Stones of the Kidney and Ureter (ReSKU) were included if they had computed tomography images available and an accompanying official radiologic report at the time of their urologist provider visit. Stone number and aggregate stone size were compared between the urologic interpretation and the corresponding radiologic reports.

RESULTS

Of 219 patients who met the inclusion criteria, concordance between urologic and radiologic assessment of aggregate stone size was higher for single stone sizing (63%) compared with multiple stones (32%). Statistical significance was found in comparing the mean difference in aggregate stone size for single and multiple stones (P <.01). Over 33% of stone-containing renal units had a radiologic report with an unclear size estimation or size discrepancy that could lead to non-guideline-driven surgical management.

CONCLUSION

Significant variation exists between urologic and radiologic computed tomography interpretations of stone burden. Urologists should personally review patient imaging when considering stone surgical management. A standardized method for measuring and reporting stone parameters is needed among urologists and radiologists.

Imaging Advances in Urolithiasis

The prevalence of urinary stones in the United States has been described as 1 in 11 persons reporting a history of stones. Imaging plays a crucial role in diagnosis, management, and follow-up for these patients and imaging technology over the last 100 years has advanced as the disease prevalence has increased. CT remains the gold standard for imaging urolithiasis and changes in this technology, with the addition of multidetector CT and dual-energy CT, as well as the changes in utilization of CT, have decreased the radiation dose encountered by patients and allowed for improved stone detection. The use of digital tomography has been introduced for follow-up of recurrent stone formers offering the potential to lower radiation exposure over the course of a patient's lifelong treatment. However, there is still a demand for improved imaging techniques to detect smaller stones and stones in larger patients at lower radiation doses as well as the continued need for the judicious use of all imaging modalities for healthcare cost containment and patient safety.

Advances in CT imaging for urolithiasis

Urolithiasis is a common disease with increasing prevalence worldwide and a lifetime-estimated recurrence risk of over 50%. Imaging plays a critical role in the initial diagnosis, follow-up and urological management of urinary tract stone disease. Unenhanced helical computed tomography (CT) is highly sensitive (>95%) and specific (>96%) in the diagnosis of urolithiasis and is the imaging investigation of choice for the initial assessment of patients with suspected urolithiasis. The emergence of multi-detector CT (MDCT) and technological innovations in CT such as dual-energy CT (DECT) has widened the scope of MDCT in the stone disease management from initial diagnosis to encompass treatment planning and monitoring of treatment success. DECT has been shown to enhance pre-treatment characterization of stone composition in comparison with conventional MDCT and is being increasingly used. Although CT-related radiation dose exposure remains a valid concern, the use of low-dose MDCT protocols and integration of newer iterative reconstruction algorithms into routine CT practice has resulted in a substantial decrease in ionizing radiation exposure. In this review article, our intent is to discuss the role of MDCT in the diagnosis and post-treatment evaluation of urolithiasis and review the impact of emerging CT technologies such as dual energy in clinical practice.

Visual search in abdominopelvic CT interpretation: accuracy and time efficiency between coronal MPR and axial images

RATIONALE AND OBJECTIVES

The objective of this study was to compare reader accuracy and time efficiency between coronal reformats of abdominopelvic computed tomography (CT) and axial images, by means of a visual search task.

MATERIALS AND METHODS

In this experimental crossover study, a novel visual search task, containing targets placed on actual CT images, was constructed to assess reader performance on both planes. Six trials were shown to participants in each plane, at a fixed time of 0.5 seconds per slice. The task was presented to 43 junior doctors. On each trial, participants were assessed for accuracy and confidence in finding the target on a five-point scale. Statistical analysis was performed using the Wilcoxon signed rank test, and Fleiss kappa.

RESULTS

Coronal images took 40% less time to view overall. No significant difference was found in reader accuracy or reader confidence between the two planes. Interrater agreement was observed as fair, across a very large number of raters (43).

CONCLUSIONS

Target identification in the coronal plane is extremely similar to the axial plane on abdominopelvic CT in this study and offers a substantial time benefit. A perceptual limit to visual processing of CT images may contribute to this similarity. Greater use of coronal reformats in day-to-day practice could substantially improve radiologist workflow.

Prospective trial of the detection of urolithiasis on ultralow dose (sub mSv) noncontrast computerized tomography: direct comparison against routine low dose reference standard

PURPOSE

In this prospective trial we compared ultralow dose computerized tomography reconstruction algorithms and routine low dose computerized tomography for detecting urolithiasis.

MATERIALS AND METHODS

A total of 48 consenting adults prospectively underwent routine low dose noncontrast computerized tomography immediately followed by an ultralow dose series targeted at a 70% to 90% reduction from the routine low dose technique (sub mSv range). Ultralow dose series were reconstructed with filtered back projection, and adaptive statistical and model based iterative reconstruction techniques. Transverse (axial) and coronal images were sequentially reviewed by 3 relatively inexperienced trainees, including a radiology resident, a urology fellow and an abdominal imaging fellow. Three experienced abdominal radiologists independently reviewed the routine low dose filtered back projection images, which served as the reference standard.

RESULTS

The mean effective dose for the ultralow dose scans was 0.91 mSv (median 0.82), representing a mean ± SD 78% ± 5% decrease compared to the routine low dose. Overall sensitivity and positive predictive value per stone for ultralow dose computerized tomography at a 4 mm threshold was 0.91 and 0.98, respectively. Sensitivity, specificity, positive and negative predictive values, and accuracy per patient were 0.87, 1.00, 1.00, 0.94 and 0.96, respectively. At a 4 mm threshold the sensitivity and positive predictive value per stone of the ultralow dose series for filtered back projection, and adaptive statistical and model based iterative reconstruction was 0.89 and 0.96, 0.91 and 0.98, and 0.93 and 1.00, respectively. Sensitivity, specificity, positive and negative predictive values, and accuracy per patient at the 4 mm threshold were 0.82, 1.00, 1.00, 0.91 and 0.94 for filtered back projection, 0.85, 1.00, 1.00, 0.93 and 0.95 for adaptive statistical iterative reconstruction, and 0.94, 1.00, 1.00, 0.97 and 0.98 for model based iterative reconstruction, respectively. Sequential review of coronal images changed the final stone reading in 13% of cases and improved diagnostic confidence in 49%.

CONCLUSIONS

At a 4 mm renal calculus size threshold ultralow dose computerized tomography is accurate for detection when referenced against routine low dose series with dose reduction to below the level of a typical 2-view plain x-ray of the kidneys, ureters and bladder. Slight differences were seen among the reconstruction algorithms. There was mild improvement with model based iterative reconstruction over filtered back projection and adaptive statistical iterative reconstruction. Coronal images improved detection and diagnostic confidence over axial images alone.

Unenhanced MDCT in suspected urolithiasis: improved stone detection and density measurements using coronal maximum-intensity-projection images

OBJECTIVE

The purpose of this study was to determine whether coronal maximum-intensity-projection (MIP) reformations improve urinary tract stone detection and density measurements compared with routine axial and coronal images.

MATERIALS AND METHODS

Forty-five consecutive patients who underwent MDCT for suspected urolithiasis were included. Two radiologists independently determined the number of stones on 5-, 3-, and 1.25-mm axial, 5- and 3-mm coronal, and 5-mm coronal MIP images. The reference standard was obtained by consensus review using all six datasets. Stone density was determined for all calculi 4 mm or larger on all datasets.

RESULTS

There were a total of 115 stones. Reader 1 identified 111 (96.5%), 112 (97.4%), 97 (84.3%), 102 (88.7%), 99 (86.1%), and 85 (73.9%) stones and reader 2 identified 105 (91.3%), 102 (88.7%), 85 (73.9%), 89 (77.4%), 89 (77.4%), and 76 (66.1%) stones on the MIP, 1.25-mm axial, 3-mm axial, 3-mm coronal, 5-mm coronal, and 5-mm axial images, respectively. Both readers identified more stones on the MIP images than on the 3- or 5-mm axial or coronal images (p < 0.0001). The mean difference in stone attenuation compared with the thin axial images was significantly less for the MIP images (44.6 HU) compared with 3-mm axial (235 HU), 3-mm coronal (309 HU), and 5-mm coronal (329.6 HU) or axial images (347.8 HU) (p < 0.0001).

CONCLUSION

Coronal MIP reformations allow more accurate identification and density measurements of urinary tract stones compared with routine axial and coronal reformations.

Current computed tomography techniques can detect duct of Bellini plugging but not Randall's plaques

OBJECTIVE

To assess the ability of noninvasive computed tomography (CT) scans to detect interstitial calcium phosphate deposits (Randall's plaques) and duct of Bellini plugs, which are possible stone precursor lesions.

METHODS

At time of percutaneous nephrolithotomy (PCNL) for stone removal, all accessible individual papillae of 105 patients were endoscopically visualized and video recorded. Image-processing software was used to estimate the percentage of papillary surface occupied by plaque or plug in each pole (upper, middle, lower). The location of stones was also recorded. A radiologist blinded to the mapping results scored presurgical (n = 98) and postsurgical (n = 105) abdominal CT scans for the presence or absence of calcification by pole.

RESULTS

The cohort was a mean age of 56 years (range, 23-84 years). Maximum papillary surface area of each area of the kidney occupied by plug correlated with CT calcifications on pre- and postprocedure images by rank sum test. However, maximum plaque surface area did not correlate with radiographic findings (P = .10-.90 for each pole by rank sum test). Sensitivity was 81% and specificity was 69% of CT to detect plugs of at least 1% of the papillary surface area.

CONCLUSION

Calcifications seen on current generation clinical CT scans correspond to ductal plugging involving at least 1% of the papillary surface area. Current clinical CT scan technology appears inadequate for detecting Randall's plaques.

Nephrolithiasis: what surgeons need to know

OBJECTIVE

In this article, we review the standard of care for imaging of nephrolithiasis as well as new technology and radiation concerns from the perspective of the urologic surgeon.

CONCLUSION

Nephrolithiasis is a common cause of morbidity with a lifetime prevalence of 5-10% worldwide. Increasingly, diagnostic evaluation and planning for medical or surgical intervention have become reliant on imaging.

Use of a Dedicated Server to Perform Coronal and Sagittal Reformations in Trauma Examinations

The purpose of this study was to evaluate the impact of implementing an automated process for generating coronal and sagittal reformatted images on radiologist workflow. When performing trauma-related CT examinations of the cervical, thoracic, and lumbar spine at our institution, technologists manually generate coronal and sagittal reconstructions at the scanner console and send these images to a picture archiving and communication system (PACS) for interpretation by radiologists and clinical viewing. Although certain PACS, thin-client three-dimensional systems, and CT scanners are capable of automatically generating reconstructed or reformatted images, the systems at our institution do not support this functionality. We have recently integrated a dedicated server that is capable of automatically generating multiplanar reformatted (MPR) images from source thin-section axial images and sending these images to PACS without requiring technologist input. This dedicated server was used to generate coronal and sagittal MPRs for trauma-related spine studies in parallel with technologist-generated coronal and sagittal reconstructions. When comparing the two methods, using the dedicated server to automatically generate reformations resulted in substantial time savings for the radiologist compared to technologist-generated reconstructions. Additionally, a survey of interpreting radiologists indicated that a significant majority preferred to view the automatically generated MPRs on PACS compared to the thin-client system, considered the image quality to be good or excellent, and believed that viewing MPRs increased diagnostic accuracy and confidence. It is expected that this automated process will significantly improve radiologist workflow with respect to image interpretation time and report turnaround time.