Identifying urethral strictures using machine learning: a proof-of-concept evaluation of convolutional neural network model.

INTRODUCTION: To evaluate urethral strictures and to determine appropriate surgical reconstructive options, retrograde urethrograms (RUG) are used. Herein, we develop a convolutional neural network (CNN)-based machine learning algorithm to characterize RUG images between those with urethral strictures and those without urethral strictures. METHODS: Following approval from institutional REB from participating institutions (The Hospital for Sick Children [Toronto, Canada], St. Luke's Medical Centre [Quezon City, Philippines], East Virginia Medical School [Norfolk, United States of America]), retrograde urethrogram images were collected and anonymized. Additional RUG images were downloaded online using web scraping method through Selenium and Python 3.8.2. A CNN with three convolutional layers and three pooling layers were built (Fig. 1). Data augmentation was applied with zoom, contrast, horizontal flip, and translation. The data were split into 90% training and 10% testing set. The model was trained with one hundred epochs. RESULTS: A total of 242 RUG images were identified. 196 were identified as strictures and 46 as normal. Following training, our model achieved accuracy of up to 92.2% with its training data set in characterizing RUG images to stricture and normal images. The validation accuracy using our testing set images showed that it was able to characterize 88.5% of the images correctly. CONCLUSION: It is feasible to use a machine learning algorithm to accurately differentiate between a stricture and normal RUG. Further development of the model with additional RUGs may allow characterization of stricture location and length to suggest optimal operative approach for repair.

Diagnostic accuracy of contrast-enhanced voiding urosonogram using second-generation contrast with harmonic imaging (CEVUS-HI) study for assessment of vesicoureteral reflux in children: a meta-analysis.

PURPOSE: To assess the diagnostic accuracy and safety of contrast-enhanced voiding urosonogram using second-generation contrast with harmonic imaging (CEVUS-HI) in detecting vesicoureteral reflux (VUR) among children. METHODS: A systematic literature search was performed in March 2018. Relevant comparative studies from Medline, EMBASE, World of Science, Scopus, CENTRAL, WHO trial registry and Clinicaltrials.gov were identified and appraised using QUADAS-2. Diagnostic accuracy parameters were determined using VCUG as the reference standard. Adverse effects related to ultrasound contrast were summarized. The heterogeneity and inter-study variability were determined. After appropriate subgroup diagnostic accuracy parameters were investigated, summarizing receiver operator characteristics was constructed using the bivariate model meta-regression to determine the area under the curve (AUC). RESULTS: A total of 12 studies with low-high risk of bias, including 1917 ureteral units from 953 patients were assessed for this meta-analysis. The included studies reported no serious adverse events associated with the ultrasound contrast. The pooled diagnostic accuracy parameters of CEVUS-HI in detecting VUR amongst children were: sensitivity 90.43 (95% CI 90.36-90.50), specificity 92.82 (95% CI 92.76-92.87), the calculated (+) likelihood-ratio 12.59 (95% CI 12.49-12.68), (-) likelihood-ratio of 0.103 (95% CI 0.102-0.104) and extrapolated pooled diagnostic odds-ratio was 122.12 (95% CI 120.75-123.49). Heterogeneity with interstudy variability was noted (p < 0.0001, I-squared > 70%). The AUC was determined to be 0.965 for VUR detection. CONCLUSIONS: The pooled diagnostic accuracy parameters from low-moderate quality of evidence have illustrated that the CEVUS-HI study has an excellent safety profile and acceptable diagnostic accuracy. It may be considered as an alternative diagnostic modality for assessment of VUR among children.

Circumcision-incision orchidopexy: A novel technique for palpable, low inguinal undescended testis.

Given that both orchidopexy and circumcision are commonly done in a single operative setting, we adopted a technique of combined orchidopexy and circumcision using a single circumcision incision. We applied this new technique to boys with palpable, low inguinal cryptorchidism. Here we describe a case series of 7 boys who underwent concurrent orchidopexy via the circumcision site. We present this novel technique and discuss our preliminary outcomes, including the anatomic basis and feasibility. The technique appears to be an alternative for concurrent circumcision and cryptorchid cases with palpable, low inguinal testes.

Use of computed tomography scout film and Hounsfield unit of computed tomography scan in predicting the radio-opacity of urinary calculi in plain kidney, ureter and bladder radiographs.

OBJECTIVE: The objective of this study is to determine the diagnostic utility of computed tomography (CT)- scout film with an optimal non-contrast helical CT scan Hounsfield unit (HU) in predicting the appearance of urinary calculus in the plain kidneys, ureter, urinary bladder (KUB)-radiograph. METHODS: A prospective cross-sectional study was executed and data were collected from June 2007 to June 2012 at a tertiary hospital. The included subjects were diagnosed to have <10mm urolithiasis with non-contrast helical CT scan and KUB X-ray, which were carried out on the same day. Both KUB radiographs and CT-scout film were read by two qualified radiologists with inter-observer standardization prior to the study. Urolithiasis characteristics such as stone location, CT attenuation value, CT-scout film and KUB radiograph appearance were recorded independently by two observers. Univariate logistic analysis with receiver operating characteristic curve was generated to determine the best cut-off HU value of urolithiases not identified in CT-scout film, but determined radio-opaque in KUB X-ray. Subsequently, its sensitivity, specificity, predictive values and likelihood ratios were calculated. Statistical significance was set at P value of 0.05 or less. RESULTS: Two hundred and three valid cases were included. 73 out of 75 CT-scout film detected urolithiasis were identified on plain radiograph and determined as radio-opaque. The determined best cut off value of HU utilized for prediction of radiographic characteristics was 630HU at which urinary calculi were not seen at CT-scout film and were KUB X-ray radio-opaque. The set HU cut-off was established of ideal accuracy with an overall sensitivity of 82.2%, specificity of 96.9% and a positive predictive value of 96.5% and negative predictive value of 83.5%. CONCLUSION: Urolithiases identified on the CT-scout film were also seen as radiopaque on the KUB radiograph while those stones not visible on the CT-scout film, but above the optimal HU cut-off value of 630 are also likely to be radiopaque.