Deep learning of renal scans in children with antenatal hydronephrosis.

INTRODUCTION: Antenatal hydronephrosis (ANH) is one of the most common anomalies identified on prenatal ultrasound, found in up to 4.5% of all pregnancies. Children with ANH are surveilled with repeated renal ultrasound and when there is high suspicion for a ureteropelvic junction obstruction on renal ultrasound, a mercaptuacetyltriglycerine (MAG3) Lasix renal scan is performed to evaluate for obstruction. However, the challenging interpretation of MAG3 renal scans places patients at risk of misdiagnosis. OBJECTIVE: Our objective was to analyze MAG3 renal scans using machine learning to predict renal complications. We hypothesized that our deep learning model would extract features from MAG3 renal scans that can predict renal complications in children with ANH. STUDY DESIGN: We performed a case-control study of MAG3 studies drawn from a population of children with ANH concerning for ureteropelvic junction obstruction evaluated at our institution from January 2009 until June of 2021. The outcome was renal complications that occur ≥6 months after an equivocal MAG-3 renal scan. We created two machine learning models: a deep learning model using the radiotracer concentration versus time data from the kidney of interest and a random forest model created using clinical data. The performance of the models was assessed using measures of diagnostic accuracy. RESULTS: We identified 152 eligible patients with available images of which 62 were cases and 90 were controls. The deep learning model predicted future renal complications with an overall accuracy of 73% (95% confidence inteveral [CI] 68-76%) and an AUC of 0.78 (95% CI 0.7, 0.84). The random forest model had an accuracy of 62% (95% CI 60-66%) and an AUC of 0.67 (95% CI. 0 64, 0.72) DISCUSSION: Our deep learning model predicted patients at high risk of developing renal complications following an equivocal renal scan and discriminate those at low risk with moderately high accuracy (73%). The deep learning model outperformed the clinical model built from clinical features classically used by urologists for surgical decision making. CONCLUSION: Our models have the potential to influence clinical decision making by providing supplemental analytical data from MAG3 scans that would not otherwise be available to urologists. Future multi-institutional retrospective and prospective trials are needed to validate our model.

Historical bibliometric analysis of the top cited articles on vesicoureteral reflux 1950-2016, and incorporation of a novel impact index.

INTRODUCTION AND OBJECTIVES: Vesicoureteral reflux (VUR) has been one of the defining conditions unique to pediatric urology since its inception. The clinical implications of this disease process depend on intrinsic patient factors such as age, genetics, epigenetics, voiding habits, anatomic anomalies, and extrinsic factors such as the pathogenicity of infectious agents. Knowledge about its natural history, the implications of conservative and surgical management, and their associated outcomes have evolved dramatically over time. This study aimed to use bibliometric analyses to summarize the evolution of VUR management over time. In order to accomplish this, the most referenced articles for VUR since 1950 were identified, and a comprehensive analysis of their impact on the management and understanding of VUR was performed by creating a novel impact index. METHODS: A reference search was carried out for indexed citations through the portal 'Science Citation Index' in the subsection 'Web of Science Core Collection' using 'vesicoureteral reflux' as a MeSH term. References were analyzed and subcategorized according to various subtopics. A unique impact index was developed to adjust the number of publications for the time since publication, in order to define the impact of the paper amongst the most frequently cited papers. Articles were analyzed and data were tabulated according to the number of citations, country and institute of origin, journal of publication, impact factor, and first authorship. RESULTS: Citation counts ranged from 43 to 510, and the mean number of citations per publication was 101.43. The most discussed topic was 'treatment'. The impact index showed that more recent publications have a higher impact. The author with the highest index impact had 271 citations in a period of 5 years. The top 150 articles were published across 23 countries, the majority being from the USA (Summary fig.). The most frequently cited institution had 12 publications. The journal with the highest publication referencing rate was the Journal of Urology. CONCLUSION: The most cited articles were valuable sources of information to describe the historical evolution of the pathophysiology and management of VUR. After adjusting for time since publication, the most recent publications (i.e. those published after 1990) had a higher impact index. Combining traditional bibliometric analysis with this novel impact index may allow researchers to optimize future literature analyses, while also assisting clinicians in understanding best practices for patient management based on the available literature.