Combining Mean and Standard Deviation of Hounsfield Unit Measurements from Preoperative CT Allows More Accurate Prediction of Urinary Stone Composition Than Mean Hounsfield Units Alone

Advancements in stone classification: unveiling the beauty of urolithiasis

PURPOSE

Urolithiasis has become increasingly prevalent, leading to higher disability-adjusted life years and deaths. Various stone classification systems have been developed to enhance the understanding of lithogenesis, aid urologists in treatment decisions, and predict recurrence risk. The aim of this manuscript is to provide an overview of different stone classification criteria.

METHODS

Two authors conducted a review of literature on studies relating to the classification of urolithiasis. A narrative synthesis for analysis of the studies was used.

RESULTS

Stones can be categorized based on anatomical position, size, medical imaging features, risk of recurrence, etiology, composition, and morphoconstitutional analysis. The first three mentioned offer a straightforward approach to stone classification, directly influencing treatment recommendations. With the routine use of CT imaging before treatment, precise details like anatomical location, stone dimensions, and Hounsfield Units can be easily determined, aiding treatment planning. In contrast, classifying stones based on risk of recurrence and etiology is more complex due to dependencies on multiple variables, including stone composition and morphology. A classification system based on morphoconstitutional analysis, which combines morphological stone appearance and chemical composition, has demonstrated its value. It allows for the rapid identification of crystalline phase principles, the detection of crystalline conversion processes, the determination of etiopathogenesis, the recognition of lithogenic processes, the assessment of crystal formation speed, related recurrence rates, and guidance for selecting appropriate treatment modalities.

CONCLUSIONS

Recognizing that no single classification system can comprehensively cover all aspects, the integration of all classification approaches is essential for tailoring urolithiasis patient-specific management.

Predicting stone composition via machine-learning models trained on intra-operative endoscopic digital images

OBJECTIVES

The aim of this study was to use deep learning (DL) of intraoperative images of urinary stones to predict the composition of urinary stones. In this way, the laser frequency and intensity can be adjusted in real time to reduce operation time and surgical trauma.

MATERIALS AND METHODS

A total of 490 patients who underwent holmium laser surgery during the two-year period from March 2021 to March 2023 and had stone analysis results were collected by the stone laboratory. A total of 1658 intraoperative stone images were obtained. The eight stone categories with the highest number of stones were selected by sorting. Single component stones include calcium oxalate monohydrate (W1), calcium oxalate dihydrate (W2), magnesium ammonium phosphate hexahydrate, apatite carbonate (CH) and anhydrous uric acid (U). Mixed stones include W2 + U, W1 + W2 and W1 + CH. All stones have intraoperative videos. More than 20 intraoperative high-resolution images of the stones, including the surface and core of the stones, were available for each patient via FFmpeg command screenshots. The deep convolutional neural network (CNN) ResNet-101 (ResNet, Microsoft) was applied to each image as a multiclass classification model.

RESULTS

The composition prediction rates for each component were as follows: calcium oxalate monohydrate 99% (n = 142), calcium oxalate dihydrate 100% (n = 29), apatite carbonate 100% (n = 131), anhydrous uric acid 98% (n = 57), W1 + W2 100% (n = 82), W1 + CH 100% ( n = 20) and W2 + U 100% (n = 24). The overall weighted recall of the cellular neural network component analysis for the entire cohort was 99%.

CONCLUSION

This preliminary study suggests that DL is a promising method for identifying urinary stone components from intraoperative endoscopic images. Compared to intraoperative identification of stone components by the human eye, DL can discriminate single and mixed stone components more accurately and quickly. At the same time, based on the training of stone images in vitro, it is closer to the clinical application of stone images in vivo. This technology can be used to identify the composition of stones in real time and to adjust the frequency and energy intensity of the holmium laser in time. The prediction of stone composition can significantly shorten the operation time, improve the efficiency of stone surgery and prevent the risk of postoperative infection.

Development and external validation of a machine learning-based model to classify uric acid stones in patients with kidney stones of Hounsfield units < 800

The correct diagnosis of uric acid (UA) stones has important clinical implications since patients with a high risk of perioperative morbidity may be spared surgical intervention and be offered alkalization therapy. We developed and validated a machine learning (ML)-based model to identify stones on computed tomography (CT) images and simultaneously classify UA stones from non-UA stones. An international, multicenter study was performed on 202 patients who received percutaneous nephrolithotomy for kidney stones with HU < 800. Data from 156 (77.2%) patients were used for model development, while data from 46 (22.8%) patients from a multinational institution were used for external validation. A total of 21,074 kidney and stone contour-annotated CT images were trained with the ResNet-18 Mask R-convolutional neural network algorithm. Finally, this model was concatenated with demographic and clinical data as a fully connected layer for stone classification. Our model was 100% sensitive in detecting kidney stones in each patient, and the delineation of kidney and stone contours was precise within clinically acceptable ranges. The development model provided an accuracy of 99.9%, with 100.0% sensitivity and 98.9% specificity, in distinguishing UA from non-UA stones. On external validation, the model performed with an accuracy of 97.1%, with 89.4% sensitivity and 98.6% specificity. SHAP plots revealed stone density, diabetes mellitus, and urinary pH as the most important features for classification. Our ML-based model accurately identified and delineated kidney stones and classified UA stones from non-UA stones with the highest predictive accuracy reported to date. Our model can be reliably used to select candidates for an earlier-directed alkalization therapy.

Analysis and prediction of postoperative recurrence of unilateral upper ureteral calculi in 243 cases (nomogram)

Objective: This study aimed to identify the risk factors for postoperative recurrence of unilateral upper ureteral calculi and develop a predictive nomogram. Patients and Methods: A retrospective analysis was conducted on 243 patients diagnosed with unilateral upper ureteral calculi who were treated at our hospital between January 1, 2016 and December 31, 2018. Patients were divided into two groups: recurrence or non-recurrence cohort. Differences in age, gender, smoking and/or drinking habit, laterality, stone diameter, ureteral stricture, stone incarceration, urinary tract infection, surgical intervention, operation time, body mass index, and metabolic syndrome were analyzed. Discrete risk factors were screened, and a nomogram was developed to predict the probability of stone recurrence. Results: The study found that the recurrence of ureteral calculi was associated with factors including stone diameter, ureteral stricture, stone incarceration, surgical intervention, operation time, metabolic syndrome, body mass index, triglycerides, diabetes, and high blood pressure (p < 0.05). Ureteral stricture, surgical intervention, metabolic syndrome, and triglycerides were found to be discrete risk factors for stone recurrence (p < 0.05). In addition, the study revealed that the stone recurrence rate of metabolic syndrome patients was significantly elevated (p < 0.05), as demonstrated by the survival curve. Lastly, using the nomogram, with an area under the curve value of 0.929, the recurrence rate of ureteral calculi was predicted. Conclusions: The study identified that preoperative ureteral stricture, laparoscopic ureterolithotomy, metabolic syndrome, and triglycerides are closely related to postoperative recurrence of ureteral calculi. The nomogram developed in this study can be used as a predictive tool for the recurrence rate of ureteral calculi.

The combination of mean and maximum Hounsfield Unit allows more accurate prediction of uric acid stones

Based on mean Hounsfield Unit (HuMean), we aimed to evaluate the additional use of standard deviation of Hounsfield Unit (HuStd), minimum Hounsfield Unit (HuMin), and maximum Hounsfield Unit (HuMax) in noncontrast computed tomography (NCCT) to evaluate uric acid (UA) stones more accurately. The data of patients who underwent the NCCT examination and infrared spectroscopy in our hospital from August 2017 to December 2021 were analyzed retrospectively. Based on CT scans, the HuMean, HuStd, HuMin, and HuMax of all patients were measured. The patients were divided into groups according to the stone composition. The attenuation value of mixed stones was in the middle of their pure stones. Except for Str, statistically significant differences between UA stones and other pure stones were observed for HuMean, HuStd, HuMin, and HuMax. A moderate correlation was found between HuMean, HuStd, HuMin, and HuMax and UA stones (r_s showed -0.585, -0.409, -0.492, and -0.577, respectively). Receiver operator characteristic (ROC) curve showed that the area under the curve (AUC) of HuMean and HuMax were higher than those of HuStd and HuMin (AUC = 0.896, AUC = 0.891 vs. AUC = 0.777, AUC = 0.833). Higher AUC (0.904), specificity (0.899) and positive predictive value (PPV) (0.712) can be obtained by combining HuMean and HuMax in the diagnosis of UA stones. In conclusion, HuMean and HuMax can better predict UA stones than HuStd and HuMin. The combined use of HuMean and HuMax can lead to higher accuracy.

Revolution spectral CT for urinary stone with a single/mixed composition in vivo: a large sample analysis

PURPOSE

To analyze various compositions of urinary stones using revolution spectral CT (rapid kV switching dual-energy CT) in vivo.

METHODS

202 patients with urinary stones underwent spectral CT before surgery. Zeff peak, overall scope and CT values were detected. Moreover, water/iodine attenuating material images were obtained. Removed stones were subjected to infrared spectroscopy after surgery. The results of infrared spectroscopy were compared with CT.

RESULTS

28 stones (14.08%) with single composition, 165 stones with two mixed compositions (81.68%), and 9 stones with three mixed compositions (4.46%) were observed. When Zeff peaks of stones with single/mixed compositions were summarized together, 146 peaks of calcium oxalate monohydrate, 119 peaks of calcium oxalate dihydrate, 55 peaks of carbapatite, 38 peaks of urate, 16 peaks of struvite, and 11 peaks of brushite were totally observed. 93.8% of calcium oxalate monohydrate had Zeff peaks between 13.3 and 14.0. 91.6% of calcium oxalate dihydrate had peaks between 12.0 and 13.3. For carbapatite, 90.9% of stones had peaks from 14.0 to 15.0. A total of 94.8% of urate had peaks between 7.0 and 11.0. 93.8% of struvite had peaks between 11.0 and 13.0, and 90.9% of brushite had peaks between 12.0 and 14.0. Moreover, densities of urate, struvite and brushite were low density in iodine-based images and high-density in water-based images.

CONCLUSION

The in-vivo analysis of spectral CT in urinary stone revealed characteristics of different compositions, especially mixed compositions. An in-vivo predictive model may be constructed to distinguish stone compositions.

A novel method for prediction of stone composition: the average and difference of Hounsfield units and their cut-off values

PURPOSE

The purpose of the study was to investigate the predictive value of stone measurements by including a novel method on non-contrast computed tomography (NCCT) images for stone composition.

METHODS

We retrospectively evaluated patients who had stone analysis, NCCT images, and underwent percutaneous nephrolithotomy between 2013 and 2016. Patient characteristics, stone measurements on NCCT images, and stone analysis results were evaluated. Hounsfield unit (HU) values (maximum (HU_max), minimum (HU_min), and average (HU_ave) of HU values) were investigated on NCCT images. HU_diff was calculated as the difference between the HU_max and the HU_min values. Patients were divided into seven stone groups and data were compared. Then patients were separately divided into two groups according to mineral complexity (mono-mineral and multi-mineral groups) and calcium-based (calcium and other stone groups) evaluation.

RESULTS

In the study, 115 patients were evaluated. Age, gender, HU_min, HU_max, and HU_ave were significantly different between the stone groups. HU_diff and HU_ave were found to be 341.5 HU (AUC = 0.719, p = 0.017) and 1051.5 HU (AUC = 0.701, p = 0.029) as cut-off, respectively. Seventy of 72 > 341.5 HU_diff patients and 64 of 67 > 1051.5 HU_ave patients had multi-mineral stones (p = 0.001, OR 9.26, and p = 0.028, OR 4.27), respectively. In multivariate analysis, > 341.5 HU_diff rate was significantly higher in multi-mineral and calcium stone groups; HU_ave was also significantly higher in the calcium stone group.

CONCLUSIONS

HU_diff and HU_ave were significant predictors of mineral complexity. HU_diff of < 341.5 HU showed 81.8% sensitivity and 67.2% specificity for identification of mono-mineral stones.

Accurately Diagnosing Uric Acid Stones from Conventional Computerized Tomography Imaging: Development and Preliminary Assessment of a Pixel Mapping Software

PURPOSE

Preoperative determination of uric acid stones from computerized tomography imaging would be of tremendous clinical use. We sought to design a software algorithm that could apply data from noncontrast computerized tomography to predict the presence of uric acid stones.

MATERIALS AND METHODS

Patients with pure uric acid and calcium oxalate stones were identified from our stone registry. Only stones greater than 4 mm which were clearly traceable from initial computerized tomography to final composition were included in analysis. A semiautomated computer algorithm was used to process image data. Average and maximum HU, eccentricity (deviation from a circle) and kurtosis (peakedness vs flatness) were automatically generated. These parameters were examined in several mathematical models to predict the presence of uric acid stones.

RESULTS

A total of 100 patients, of whom 52 had calcium oxalate and 48 had uric acid stones, were included in the final analysis. Uric acid stones were significantly larger (12.2 vs 9.0 mm, p = 0.03) but calcium oxalate stones had higher mean attenuation (457 vs 315 HU, p = 0.001) and maximum attenuation (918 vs 553 HU, p <0.001). Kurtosis was significantly higher in each axis for calcium oxalate stones (each p <0.001). A composite algorithm using attenuation distribution pattern, average attenuation and stone size had overall 89% sensitivity, 91% specificity, 91% positive predictive value and 89% negative predictive value to predict uric acid stones.

CONCLUSIONS

A combination of stone size, attenuation intensity and attenuation pattern from conventional computerized tomography can distinguish uric acid stones from calcium oxalate stones with high sensitivity and specificity.