ESPU Meeting on Saturday 6, September 2025, 08:00 - 08:50
08:00 - 08:03
S35-1 (OP)
Faruk ARSLAN 1, Omer ALGORABI 2, Yusuf Sait TURKAN 2, Ersin NAMLI 2, Onur Can OZKAN 3, Tufan TARCAN 3, Selcuk YUCEL 3 and Cagri Akin SEKERCI 3
1) Marmara University Pendik Training and Research Hospital, Urology, Istanbul, TÜRKIYE - 2) Istanbul University, Cerrahpaşa, Faculty of Engineering, Department of Industrial Engineering, Istanbul, TÜRKIYE - 3) Marmara University Pendik Training and Research Hospital, Paediatric Urology, Istanbul, TÜRKIYE
PURPOSE
In the literature, the studies have highlighted a weak agreement among experts in interpreting uroflowmetry patterns. The study aims to assess the impact of machine learning (ML) models on the interpretation of voiding patterns.
MATERIAL AND METHODS
The study included uroflowmetries of children aged 4–17 years with LUTS. Uroflowmetry patterns were independently interpreted by three pediatric urologist. Discrepancies in interpretations were jointly re-evaluated, and a consensus was reached. Voiding volume, duration, and flow rates were converted into numerical data. Eighty percent of the dataset was used as training data for ML, while there maining 20% was reserved for testing. Five different ML models were employed for classification: Decision Tree, Random Forest, CatBoost, XGBoost, and LightGBM. The models that most accurately identified each pattern were determined.
RESULTS
A total of 500 (221 boys (44.2%) and 279 girls (55.8%); 9.17 ± 3.41 years) uroflowmetries were included. In the initial assessment, 311 tests (62.2%) were interpreted similarly by the observers, while 189 tests (37.8%) were interpreted differently by at least one observer (Fleiss'Kappa=0.608). Of the samples used for ML training, 253 (50.6%) exhibited a bell-shaped pattern, 52 (10.4%) tower, 103 (20.6%) staccato, 40 (8%) intermittent, and 52 (10.4%) plateau. The highest accuracy was achieved with XGBoost (85.00%±2.90), while the lowest accuracy was observed with the Decision Tree (81.80%±1.47)(Table 1).
|
Voiding Patterns |
Highest Accuracy |
Accuracy rate (%) |
Lowest Accuracy |
Accuracy rate (%) |
|
Bell Shaped |
XGBoost |
90,91 |
Decision Tree |
85,77 |
|
Tower |
Random Forest |
73,08 |
Decision TreeXGBoostLightGBM |
63,46 |
|
Staccato |
CatBoost |
83,50 |
LightGBM |
79,61 |
|
Intermittent |
XGBoost LightGBM |
100 |
Random Forest |
95 |
|
Plateau |
XGBoost |
71,15 |
LightGBM |
61,54 |
CONCLUSIONS
High accuracy rates were observed for ML models in terms of interpreting uroflowmetry patterns in children. We believe that artificial intelligence models could contribute to the interpretation and standardization of voiding patterns in the future.
08:03 - 08:06
S35-2 (OP)
Wouter VAN DORT 1, Peter ROSIER 2, Ruud WORTEL 1, Rogier SCHROEDER 1 and Laetitia DE KORT 2
1) UMC Utrecht, Pediatric Urology, Utrecht, NETHERLANDS - 2) UMC Utrecht, Urology, Utrecht, NETHERLANDS
PURPOSE
Uroflowmetry is a widely used diagnostic to assess the physiology of the voiding. Unfortunately, this tool cannot be used in very young children, as they are not able to sit on a uroflowmetry toilet. This also reduces the diagnostic value of the urodynamic pressure-flow study in assessing bladder outflow obstruction and bladder contractility in these children. Therefore, other methods for the derivation of the uroflowmetry curve could be helpful. We explored a method to calculate the uroflowmetry curve using standard care video urodynamic studies (VUDS).
MATERIAL AND METHODS
VUDS of 50 patients (48% boy, median age 10 years), able to void on a uroflowmetry toilet were retrospectively included, resulting in 1147 X-ray images. The area of the bladder was manually segmented, and the consecutive images were converted to a uroflowmetry curve, called videoflow. This manual method was enhanced with an AI algorithm, resulting in an automated derivation of the videoflow. Differences between the videoflow and standard uroflowmetry were assessed.
RESULTS
An excellent cross-correlation of 0.98 of the videoflow with the standard uroflowmetry was found. The Qmax was not significantly different (p=0.874), with a mean difference of only 0.4%. The mean absolute deviation in Qmax was 1.3ml/s in the boys and 2.2ml/s in the girls. The AI model performed excellent, with only a 2% deviation from the manual segmentation.
CONCLUSIONS
The videoflow is an accurate method to derive the uroflowmetry curve from a VUDS. If validated in very young children, videoflow will be a potential new tool to assess bladder outflow obstruction in this group. The method could be extended to voiding cystographies, enlarging it’s potential for clinical use.
08:06 - 08:09
S35-3 (OP)
Naila MUSTAFAYEVA 1, Osman Hakan KOCAMAN 2, Erbug KESKIN 3, Feryal Gun SOYSAL 2 and Basak ERGINEL 2
1) Ege Hospital, Pediatic surgery, Baku, AZERBAIJAN - 2) Istanbul Medical Faculty, Pediatic surgery, Istanbul, TÜRKIYE - 3) Istanbul Medical Faculty, Istanbul, TÜRKIYE
PURPOSE
Uroflowmetry a non-invasive, simple, and low-cost method that can be used in the diagnosis and treatment follow-up of voiding dysfunction. There are difficulties of performing uroflow in the pediatric population. Audiflow or sonouroflow involves analyzing the sound produced by urine hitting the water in the toilet during urination to obtain a uroflow curve.
MATERIAL AND METHODS
In this study, we compared these two methods by recording the sound of urination during the conventional uroflow procedure in male children who had no lower urinary system pathology. The sound recordings were analyzed in the open-source Sonic Visualiser v4.5.2 program, and sound curves were created using the Loudness transformation, which is available under the GNU general license. Then, the curve created was digitally superimposed on the conventional UFM curve using the GIMP program. Thus, a single graph was created that we could evaluate the similarity between the noise graph of the sound recordings and the flow curve of the UFM.
RESULTS
Average age of 7.36 (ranging from 4-14 years), and an average height of 119.4 cm (ranging from 110-165 cm). Although UFM was performed with a urination sound recording for a total of 20 patients, 8 records were excluded from the study due to artifacts caused by unexpected external noises. It was observed that the audiflow curve was similar to the conventional uroflow curve, and the urination time also correlated with the conventional method.
CONCLUSIONS
With its advantages such as being simple, easy, and frequently repeatable, audiflow could be an alternative to conventional UFM.
08:09 - 08:12
S35-4 (OP)
Anil KUMAR 1, Selvin Theodore Jayanth DANIEL EZHILARASU 2, Sudhindra J 3, Partho MUKHERJEE 3 and Chandrasingh J 4
1) ZYMUS HOSPITAL, UROLOGY, Bengaluru, INDIA - 2) Christian Medical College and Vellore, Pediatric Urology, Ranipet, INDIA - 3) Christian Medical College, Vellore, Urology, Ranipet, INDIA - 4) Christian Medical College, Vellore, Pediatric Urology, Ranipet, INDIA
PURPOSE
To study the reproducibility of uroflowmetry in children aged 5-15 years and to establish the age at which uroflowmetry is representative.
MATERIAL AND METHODS
Of 291 boys screened, Uroflowmetry was done in 247 eligible boys. The flow was repeated at 2 weeks. A total of 227 children were included for analysis. 20 were excluded due to a voided volume less than 50ml and/or an interrupted pattern of flow. The maximum flow rate (Qmax), average flow rate (Qavg), voided volume (vv), time to maximum flow, flow time and voiding time was recorded. The reproducibility of all parameters was assessed using Inter-correlation coefficient (ICC) and correlation using Pearson’s coefficient.
RESULTS
Qmax, Qavg showed excellent concordance in boys of 5-15 years [ICC >0.6], implying the flows were reproducible. VV showed excellent concordance in boys of 8-15 years [ICC >0.6], but showed only moderate correlation in 5-7 years [ICC 0.4-0.6] probably due to the wide variation in the voided volumes. Qmax and Qavg showed significant correlation, in all age groups, with voided volume, in first and second flow [Pearson’s coefficient: 0.49 & 0.53 for Qmax1 and Qmax2; 0.48 & 0.51 Qavg1 and Qavg2 respectively], implying reproducibility of flow in this age group.
CONCLUSIONS
Uroflowmetry is reproducible in children aged 5-15 years. Qmax and Qavg showed excellent reproducibility and voided volume showed good reproducibility. Qmax shows excellent reproducibility even in the smallest of the age groups studied [5-7 years]
08:27 - 08:30
S35-5 (OP)
Mallory ROWAN 1, Jessica LEWIS 1, Matthew PARRETT 1, Nathalia AMADO 2, Linda BAKER 2, Ajit CHAUDHARI 3 and Kristen TULCHIN-FRANCIS 1
1) Nationwide Children's Hospital, Orthopedics, Columbus, USA - 2) Nationwide Children's Hospital, Kidney and Urinary Tract Center, New Albany, USA - 3) Ohio State, School of Health and Rehabilitation Sciences, Columbus, USA
PURPOSE
Prune belly syndrome (PBS) has a broad phenotypic spectrum of abdominal wall weakness/laxity secondary to deficient/absent abdominal wall skeletal musculature. Core musculature plays crucial roles in Valsalva maneuvers and balance by providing trunk support, stability and mobility. This study assesses balance and postural control in PBS to enhance their functional outcomes and quality of life (QOL).
MATERIAL AND METHODS
Utilizing 3-D motion capture and force plate technology (3DMCT), the sharpened Romberg (sRom), Pediatric Reach Test (PRT), and Timed Up and Go (TUG) tests were used to assess center of pressure (CoP) and balance. Non-parametric data was analyzed using a Mann-Whitney U test and Spearman’s rank-order correlation coefficient.
RESULTS
Typically developing controls (n=14;age 9-35yrs;71% male) and PBS (n=7;age 7-34yrs;86% male) were studied, with no age or sex differences (p=0.681, 0.489). Abdominoplasty status included static (n=2), dynamic (n=1), no surgery (n=3), and unknown (n=1). When compared to controls, PBS subjects had 1)greater tandem stance CoP medial-lateral and anterior-posterior range (p<0.01, p=0.037) and ellipse area (p=0.037) during sRom demonstrating instabilities affecting real-world function, 2)normal dynamic balance, 3)longer TUG completion time (p=0.038) suggesting decreased balance and mobility, and 4)no differences in spinal range of motion across trunk, pelvis, thoracic, or lumbar segments during sitting and standing transitions, suggesting compensations in trunk muscle activation.
CONCLUSIONS
3DMCT quantitatively measures trunk/abdominal wall function, demonstrating the first objective evidence of postural control and balance deficits in PBS. 3DMCT warrants further study to assist judging the need for and outcomes of PBS abdominoplasty surgery and to maximize targeted rehabilitation and QOL.
08:30 - 08:33
S35-6 (OP)
Priscila ASCAR, Mariana BARCIA, Eduardo HATANAKA, Camila GENZANI, Vanessa SCARANTI, Marcelo ROMBALDI, Luciana FELTRAN, Maria Fernanda CAMARGO, Jovelino LEÃO and Paulo NOGUEIRA
Samaritano Hospital, Pediatric Renal Transplantation, São Paulo, BRAZIL
PURPOSE
Prune Belly Syndrome is a rare malformation characterized by a wide spectrum of phenotypic presentations. Many of these patients will require renal replacement therapy including renal transplantation. To categorize the manifestations of the syndrome in the affected organs, the RUBACE scoring system was created (Wong et al. BJU Int 2019; 123(1): 130-139), which takes into account, in addition to the classic triad (abdominal muscle deficiency, urinary tract dilation and cryptorchidism), extra-genitourinary anomalies.
MATERIAL AND METHODS
Retrospective analysis of pediatric patients from a single institution who received their first renal transplant from 2009 to 2021. From 500 transplants, 22 patients had Prune Belly Syndrome . Independent variables evaluated were sex, age at transplant, weight in kilograms and RUBACE profile at admission (R: renal; U: ureter; B: bladder; A: abdominal; C: cryptorchidism; E: extra-genitourinary). Outcome variables were graft loss and death. Qualitative variables were described as frequencies and continuous variables as medians and interquartile range (IQR). Mann-Whitney U test was performed to compare the results, considering p = 5%.
RESULTS
Two Prune Belly patients died after transplant. RUBACE profile was superior (22±22-23) when compared to patients who survived (16±16-19), with p=0.05. This trend was not observed when evaluating renal graft loss (p=0.13).
CONCLUSIONS
Renal transplantation in children with Prune Belly Syndrome is unique and has specific challenges. The RUBACE score is discriminatory for patient survival, but not for graft survival.
08:33 - 08:36
S35-7 (OP)
Bilal CETIN 1, Emre Burak ŞAHİNLER 2, İsmai̇l SELVİ 3, Gözde ERCAN 4, Pinar EKER 5, M.İrfan DÖNMEZ 3 and Orhun SİNANOĞLU 2
1) KARTAL DR. LUTFI KIRDAR CITY HOSPITAL, Pediatric Urology, Istanbul, TÜRKIYE - 2) Sancaktepe Şehit Prof.Dr. İlhan Varank Training and Research Hospital, Urology, İstanbul, TÜRKIYE - 3) İstanbul University İstanbul Faculty of Medicine, Division of Pediatric Urology, Department of Urology, İstanbul, TÜRKIYE - 4) Sancaktepe Şehit Prof.Dr. İlhan Varank Training and Research Hospital, Department of Pediatrics, İstanbul, TÜRKIYE - 5) Maltepe University Faculty of Medicine, Department of Biochemistry, Maltepe, TÜRKIYE
PURPOSE
There are several non-homogeneous studies with conflicting results regarding the relationship between LUTS and vitamin D. This study aimed to explore the relationship between blood vitamin D levels and LUTS in children.
MATERIAL AND METHODS
This prospective descriptive study included children aged 5-18 with LUTS (diagnosed as per DVSS), while healthy children of the same age range from a well-child clinic served as controls.Blood samples were obtained to measure serum vitamin D, calcium, phosphorus, and magnesium levels. Patients completed a two-day bladder diary and were assessed for constipation using the Rome IV criteria and Bristol stool scale. Also, height, weight, and body mass index (BMI) were recorded.Bladder diary data were used to calculate minimum, median, and maximum bladder capacities, as well as voiding frequency. Urgency and urinary incontinence episodes were recorded. DVSS results, including the final quality-of-life question, were documented.
RESULTS
The study included 136 children with LUTS and 108 controls. Median vitamin D levels were similar between the LUTS (17.15 ng/mL [IQR 12.53-22.0]) and control groups (16.60 ng/mL [IQR 12.23-21.28], p=0.148).No significant correlations were identified between vitamin D levels and bladder capacity, voiding frequency, wet nights, or DVSS scores. However, children with ≥8 voids/day had slightly higher vitamin D levels than those with ≤7 voids/day (18.56 ng/mL [IQR 15.10-23.65] vs. 17 ng/mL [IQR 12-21], p=0.037).
CONCLUSIONS
Our study showed that vitamin D levels are comparable in healthy children and those with LUTS, thus, there seems to be no relationship between vitamin D and LUTS.