Kidney stone volume estimation from computerized tomography images using a model based method of correcting for the point spread function

How to estimate stone volume and its use in stone surgery: a comprehensive review

OBJECTIVE

Current interventional guidelines refer to the cumulative stone diameter to choose the appropriate surgical modality (ureteroscopy [URS], extracorporeal shockwave lithotripsy [ESWL] and percutaneous nephrolithotomy [PCNL]). The stone volume (SV) has been introduced recently, to better estimate the stone burden. This review aimed to summarize the available methods to evaluate the SV and its use in urolithiasis treatment.

MATERIAL AND METHODS

A comprehensive review of the literature was performed in December 2022 by searching Embase, Cochrane and Pubmed databases. Articles were considered eligible if they described SV measurement or the stone free rate after different treatment modalities (SWL, URS, PCNL) or spontaneous passage, based on SV measurement. Two reviewers independently assessed the eligibility and the quality of the articles and performed the data extraction.

RESULTS

In total, 28 studies were included. All studies used different measurement techniques for stone volume. The automated volume measurement appeared to be more precise than the calculated volume. In vitro studies showed that the automated volume measurement was closer to actual stone volume, with a lower inter-observer variability. Regarding URS, stone volume was found to be more predictive of stone free rates as compared to maximum stone diameter or cumulative diameter for stones >20 mm. This was not the case for PCNL and SWL.

CONCLUSIONS

Stone volume estimation is feasible, manually or automatically and is likely a better representation of the actual stone burden. While for larger stones treated by retrograde intrarenal surgery, stone volume appears to be a better predictor of SFR, the superiority of stone volume throughout all stone burdens and for all stone treatments, remains to be proven. Automated volume acquisition is more precise and reproducible than calculated volume.

Automated Machine Learning Segmentation and Measurement of Urinary Stones on CT Scan

OBJECTIVES

To evaluate the performance of an engineered machine learning algorithm to identify kidney stones and measure stone characteristics without the need for human input.

METHODS

We performed a cross-sectional study of 94 children and adults who had kidney stones identified on non-contrast CT. A previously developed deep learning algorithm was trained to segment renal anatomy and kidney stones and to measure stone features. The performance and speed of the algorithm to measure renal anatomy and kidney stone features were compared to the current gold standard of human measurement performed by 3 independent reviewers.

RESULTS

The algorithm was 100% sensitive and 100% specific in detecting individual kidney stones. The mean stone volume segmented by the algorithm was smaller than that of human reviewers and had moderate overlap (Dice score: 0.66). There was substantial variation between human reviewers in total segmented stone volume (Jaccard score: 0.17) and volume of the single largest stone (Jaccard score: 0.33). Stone segmentations performed by the machine learning algorithm more precisely approximated stone borders than those performed by human reviewers on qualitative assessment.

CONCLUSION

An engineered machine learning algorithm can identify and characterize stones more accurately and reliably than humans, which has the potential to improve the precision and efficiency of assessing kidney stone burden.

Multi-Institutional Prospective Randomized Control Trial of Novel Intracorporeal Lithotripters: ShockPulse-SE vs Trilogy Trial

Introduction: Currently, there are multiple intracorporeal lithotripters available for use in percutaneous nephrolithotomy (PCNL). This study aimed to evaluate the efficiency of two novel lithotripters: Trilogy and ShockPulse-SE. Materials and Methods: This is a prospective multi-institutional randomized trial comparing outcomes of PCNL using two novel lithotripters between February 2019 and June 2020. The study assessed objective measures of stone clearance time, stone clearance rate, device malfunction, stone-free rates, and complications. Device assessment was provided through immediate postoperative survey by primary surgeons. Results: There were 100 standard PCNLs completed using either a Trilogy or ShockPulse-SE lithotrite. Using quantitative Stone Analysis Software to estimate stone volume, the mean stone volume was calculated at 4.18 ± 4.79 and 3.86 ± 3.43 cm3 for the Trilogy and ShockPulse-SE groups, respectively. Stone clearance rates were found to be 1.22 ± 1.67 and 0.77 ± 0.68 cm3/min for Trilogy vs ShockPulse-SE (p = 0.0542). When comparing Trilogy to ShockPulse-SE in a multivariate analysis, total operative room time (104.4 ± 48.2 minutes vs 121.1 ± 59.2 minutes p = 0.126), rates of secondary procedures (17.65% vs 40.81%, p = 0.005), and device malfunctions (1.96% vs 34.69%, p < 0.001) were less, respectively. There was no difference in final stone-free rates between devices. Conclusion: Both the Trilogy and ShockPulse-SE lithotripters are highly efficient at removing large renal stones. In this study, we noted differences between the two devices including fewer device malfunctions when Trilogy device was utilized. Clinical Trial ID number: NCT03959683.

Automated radiomic analysis of CT images to predict likelihood of spontaneous passage of symptomatic renal stones

PURPOSE

To evaluate the ability of a semi-automated radiomic analysis software in predicting the likelihood of spontaneous passage of urinary stones compared with manual measurements.

METHODS

Symptomatic patients visiting the emergency department with suspected stones in either kidney or ureters who underwent a CT scan were included. Patients were followed for up to 6 months for the outcome of a trial of passage. Maximum stone diameters in axial and coronal images were measured manually. Stone length, width, height, max diameter, volume, the mean and standard deviation of the Hounsfield units, and morphologic features were also measured using automated radiomic analysis software. Multivariate models were developed using these data to predict subsequent spontaneous stone passage, with results expressed as the area under a receiver operating curve (AUC).

RESULTS

One hundred eighty-four patients (69 females) with a median age of 56 years were included. Spontaneous stone passage occurred in 114 patients (62%). Univariate analysis demonstrated an AUC of 0.83 and 0.82 for the maximum stone diameter determined manually in the axial and coronal planes, respectively. Multivariate models demonstrated an AUC of 0.82 for a model including manual measurement of maximum stone diameter in axial and coronal planes. The same AUC was found for a model including automatic measurement of maximum height and diameter of the stone. Further addition of morphological parameters measured automatically did not increase AUC beyond 0.83.

CONCLUSION

The semi-automated radiomic analysis of urinary stones shows similar accuracy compared with manual measurements for predicting urinary stone passage. Further studies are needed to predict clinical impacts of reporting the likelihood of urinary stone passage and improving inter-observer variation using automatic radiomic analysis software.

Automatic Detection and Scoring of Kidney Stones on Noncontrast CT Images Using S.T.O.N.E. Nephrolithometry: Combined Deep Learning and Thresholding Methods

PURPOSE

To develop and validate a deep learning and thresholding-based model for automatic kidney stone detection and scoring according to S.T.O.N.E. nephrolithometry.

PROCEDURES

Abdominal noncontrast computed tomography (NCCT) images were retrospectively archived from February 2018 to April 2019 for three parts: a segmentation dataset (n = 167), a hydronephrosis classification dataset (n = 282), and test dataset (n = 117). The model consisted of four steps. First, the 3D U-Nets for kidney and renal sinus segmentation were developed. Second, the deep 3D dual-path networks for hydronephrosis grading were developed. Third, the thresholding methods were used to detect and segment stones in the renal sinus region. The stone size, CT attenuation, and tract length were calculated from the segmented stone region. Fourth, the stone's location was determined. The stone detection performance was estimated with sensitivity and positive predictive value (PPV). The hydronephrosis grading and stone size, tract length, number of involved calyces, and essence grading were estimated with the area under the curve (AUC) method and linear-weighted κ statistics, respectively.

RESULTS

The stone detection algorithm reached a sensitivity of 95.9 % (236/246) and a PPV of 98.7 % (236/239). The hydronephrosis classification algorithm achieved an AUC of 0.97. The scoring model results showed good agreement with radiologist results for the stone size, tract length, number of involved calyces, and essence grading (κ = 0.95, 95 % confidence interval [CI]: 0.92, 0.98; κ = 0.97, 95 % CI: 0.95, 1.00; κ = 0.95, 95 % CI: 0.92, 0.98; and κ = 0.97, 95 % CI: 0.94, 1.00), respectively.

CONCLUSIONS

The scoring model was constructed that can automatically detect and score stones in NCCT images.

Characterization of Urinary Stone Composition by Use of Whole-body, Photon-counting Detector CT

RATIONAL AND OBJECTIVES

This study aims to investigate the performance of a whole-body, photon-counting detector (PCD) computed tomography (CT) system in differentiating urinary stone composition.

MATERIALS AND METHODS

Eighty-seven human urinary stones with pure mineral composition were placed in four anthropomorphic water phantoms (35-50 cm lateral dimension) and scanned on a PCD-CT system at 100, 120, and 140 kV. For each phantom size, tube current was selected to match CTDI_vol (volume CT dose index) to our clinical practice. Energy thresholds at [25, 65], [25, 70], and [25, 75] keV for 100, 120, and 140 kV, respectively, were used to generate dual-energy images. Each stone was automatically segmented using in-house software; CT number ratios were calculated and used to differentiate stone types in a receiver operating characteristic (ROC) analysis. A comparison with second- and third-generation dual-source, dual-energy CT scanners with conventional energy integrating detectors (EIDs) was performed under matching conditions.

RESULTS

For all investigated settings and smaller phantoms, perfect separation between uric acid and non-uric acid stones was achieved (area under the ROC curve [AUC] = 1). For smaller phantoms, performance in differentiation of calcium oxalate and apatite stones was also similar between the three scanners: for the 35-cm phantom size, AUC values of 0.76, 0.79, and 0.80 were recorded for the second- and third-generation EID-CT and for the PCD-CT, respectively. For larger phantoms, PCD-CT and the third-generation EID-CT outperformed the second-generation EID-CT for both differentiation tasks: for a 50-cm phantom size and a uric acid/non-uric acid differentiating task, AUC values of 0.63, 0.95, and 0.99 were recorded for the second- and third-generation EID-CT and for the PCD-CT, respectively.

CONCLUSION

PCD-CT provides comparable performance to state-of-the-art EID-CT in differentiating urinary stone composition.

Shock Wave Lithotripsy is More Effective for Residual Fragments after Percutaneous Nephrolithotomy than for Primary Stones of the Same Size: A Matched Pair Cohort Study

Aims

To compare the outcome of shock wave lithotripsy (SWL) on post-percutaneous nephrolithotomy (PCNL) residual fragments (RFs) versus primary stones of the same size through a matched pair study.

Methods

Patients with a single 5-15 mm fragment 3 months after PCNL were enrolled (study group n = 59). The control group (n = 67) consisted of all adult patients with a single 5-15 mm renal stone.

Results

The success rate of SWL was significantly higher in the study group (81.4 vs. 59.7%; p = 0.008; OR: 2.95). With a cutoff point of Hounsfield units (HU) 750: the success rate was significantly lower in patients with a stone HU ≥ 750 (OR: 3.488). This HU cutoff value had no effect on the outcome of SWL in patients with post-PCNL RF (p = 0.14). On the other hand, the outcome of SWL was significantly more favorable in control group when HU < 750 (p = 0.02).

Conclusion

The success rate of SWL was 2.95-fold higher for post-PCNL RFs than in a stone burden-matched control group. The likelihood of stone clearance after SWL was 3.488-fold greater when HU was less than 750. This effect of HU was more prominent in patients receiving SWL for their primary stones while SWL was evenly effective on post PCNL RFs with different HUs.

Consistency of Renal Stone Volume Measurements Across CT Scanner Model and Reconstruction Algorithm Configurations

OBJECTIVE

The objective of this prospective study is to evaluate the consistency of renal stone volume estimation using dual-energy CT across scanner model and reconstruction algorithm configurations.

SUBJECTS AND METHODS

Patients underwent scanning with routine kidney stone composition protocols on both second- and third-generation dual-source CT scanners. Images were reconstructed using filtered back projection and iterative reconstruction (IR). In addition, a modified IR kernel on the third-generation CT scanner was evaluated. Individual kidney stone volumes were determined and compared.

RESULTS

No significant difference was noted in measured volumes between filtered back-projection data, IR data from the second-generation scanner, and the modified IR kernel data (p > 0.05). The third-generation commercially available IR kernel yielded lower volumes than did the other configurations (p < 0.0001).

CONCLUSION

With the use of a modified kernel for the third-generation scanner, patients being monitored for changes in kidney stone volume can undergo scanning performed with second- or third-generation dual-energy CT scanners, and the images obtained can be reconstructed with either filtered back projection or IR without the introduction of bias into kidney stone volume measurements.

Quantitative Prediction of Stone Fragility From Routine Dual Energy CT: Ex vivo proof of Feasibility

RATIONALE AND OBJECTIVES

Previous studies have demonstrated a qualitative relationship between stone fragility and internal stone morphology. The goal of this study was to quantify morphologic features from dual-energy computed tomography (CT) images and assess their relationship to stone fragility.

MATERIALS AND METHODS

Thirty-three calcified urinary stones were scanned with micro-CT. Next, they were placed within torso-shaped water phantoms and scanned with the dual-energy CT stone composition protocol in routine use at our institution. Mixed low- and high-energy images were used to measure volume, surface roughness, and 12 metrics describing internal morphology for each stone. The ratios of low- to high-energy CT numbers were also measured. Subsequent to imaging, stone fragility was measured by disintegrating each stone in a controlled ex vivo experiment using an ultrasonic lithotripter and recording the time to comminution. A multivariable linear regression model was developed to predict time to comminution.

RESULTS

The average stone volume was 300 mm³ (range: 134-674 mm³). The average comminution time measured ex vivo was 32 seconds (range: 7-115 seconds). Stone volume, dual-energy CT number ratio, and surface roughness were found to have the best combined predictive ability to estimate comminution time (adjusted R² = 0.58). The predictive ability of mixed dual-energy CT images, without use of the dual-energy CT number ratio, to estimate comminution time was slightly inferior, with an adjusted R² of 0.54.

CONCLUSIONS

Dual-energy CT number ratios, volume, and morphologic metrics may provide a method for predicting stone fragility, as measured by time to comminution from ultrasonic lithotripsy.

Dual-Energy CT for Quantification of Urinary Stone Composition in Mixed Stones: A Phantom Study

OBJECTIVE

The purpose of this study was to assess the feasibility of using dual-energy CT to accurately quantify uric acid and non-uric acid components in urinary stones of mixed composition.

MATERIALS AND METHODS

A total of 24 urinary stones were analyzed with micro CT to serve as the reference standard for uric acid and non-uric acid composition. These stones were placed in water phantoms to simulate body attenuation of slim to obese adults and scanned with a third-generation dual-source CT scanner by use of dual-energy modes adaptively selected on the basis of phantom size. CT number ratio, which is distinct for different materials, was calculated for each pixel of the stones. Each pixel was then classified as uric acid and non-uric acid by comparison of the CT number ratio with preset thresholds ranging from 1.10 to 1.70. Minimal, maximal, and root-mean-square errors were calculated by comparing composition with the reference standard, and the threshold with the minimal root-mean-square error was determined. A paired t test was performed to compare the stone composition determined with dual-energy CT with the reference standard obtained with micro CT.

RESULTS

The optimal CT number ratio threshold ranged from 1.27 to 1.55, dependent on phantom size. The root-mean-square error ranged from 9.60% to 12.87% across all phantom sizes. Minimal absolute error ranged from 0.04% to 1.24% and maximal absolute error from 22.05% to 35.46%. Dual-energy CT and the reference micro CT did not differ significantly on uric acid and non-uric acid composition (paired t test, p = 0.20-0.96).

CONCLUSION

Accurate quantification of uric acid and non-uric acid composition in mixed stones is possible with dual-energy CT.

Tools to improve the accuracy of kidney stone sizing with ultrasound

PURPOSE

Ultrasound (US) overestimates stone size when compared with CT. The purpose of this work was to evaluate the overestimation of stone size with US in an in vitro water bath model and investigate methods to reduce overestimation.

MATERIALS AND METHODS

Ten human stones (3-12 mm) were measured using B-mode (brightness mode) US by a sonographer blinded to the true stone size. Images were captured and compared using both a commercial US machine and software-based research US device. Image gain was adjusted between moderate and high stone intensities, and the transducer-to-stone depth was varied from 6 to 10 cm. A computerized stone-sizing program was developed to outline the stone width based on a grayscale intensity threshold.

RESULTS

Overestimation with the commercial device increased with both gain and depth. Average overestimation at moderate and high gain was 1.9±0.8 and 2.1±0.9 mm, respectively (p=0.6). Overestimation increased an average of 22% with an every 2-cm increase in depth (p=0.02). Overestimation using the research device was 1.5±0.9 mm and did not vary with depth (p=0.28). Overestimation could be reduced to 0.02±1.1 mm (p<0.001) with the computerized stone-sizing program. However, a standardized threshold consistent across depth, system, or system settings could not be resolved.

CONCLUSION

Stone size is consistently overestimated with US. Overestimation increased with increasing depth and gain using the commercial machine. Overestimation was reduced and did not vary with depth, using the software-based US device. The computerized stone-sizing program shows the potential to reduce overestimation by implementing a grayscale intensity threshold for defining the stone size. More work is needed to standardize the approach, but if successful, such an approach could significantly improve stone-sizing accuracy and lead to automation of stone sizing.

Seoul National University Renal Stone Complexity Score for Predicting Stone-Free Rate after Percutaneous Nephrolithotomy

OBJECTIVES

Currently, no standardized method is available to predict success rate after percutaneous nephrolithotomy. We devised and validated the Seoul National University Renal Stone Complexity (S-ReSC) scoring system for predicting the stone-free rate after single-tract percutaneous nephrolithotomy (sPCNL).

PATIENTS AND METHODS

The data of 155 consecutive patients who underwent sPCNL were retrospectively analyzed. Preoperative computed tomography images were reviewed. The S-ReSC score was assigned from 1 to 9 based on the number of sites involved in the renal pelvis (#1), superior and inferior major calyceal groups (#2-3), and anterior and posterior minor calyceal groups of the superior (#4-5), middle (#6-7), and inferior calyx (#8-9). The inter- and intra-observer agreements were accessed using the weighted kappa (κ). The stone-free rate and complication rate were evaluated according to the S-ReSC score. The predictive accuracy of the S-ReSC score was assessed using the area under the receiver operating characteristic curve (AUC).

RESULTS

The overall SFR was 72.3%. The mean S-ReSC score was 3.15±2.1. The weighted kappas for the inter- and intra-observer agreements were 0.832 and 0.982, respectively. The SFRs in low (1 and 2), medium (3 and 4), and high (5 or higher) S-ReSC scores were 96.0%, 69.0%, and 28.9%, respectively (p<0.001). The predictive accuracy was very high (AUC 0.860). After adjusting for other variables, the S-ReSC score was still a significant predictor of the SFR by multiple logistic regression. The complication rates were increased to low (18.7%), medium (28.6%), and high (34.2%) (p = 0.166).

CONCLUSIONS

The S-ReSC scoring system is easy to use and reproducible. This score accurately predicts the stone-free rate after sPCNL. Furthermore, this score represents the complexity of surgery.