Uroflowmetry is a standard, non-invasive method for assessing urinary flow patterns. However, its reliance on specialized equipment and in-clinic visits limits accessibility, especially for children and individuals in remote or underserved areas. This study aimed to assess the reliability and validity of video-based voiding tests compared to traditional uroflowmetry and explore the potential of integrating machine learning to automate diagnostics.

This prospective study included 120 male children, all toilet-trained and referred to our department between February and June 2024. None of the participants had undergone any prior urological interventions. During the first OPD visit, micturition videos of the children were recorded using a smartphone from side and top angles as they voided into a uroflowmetry (UFM) machine. To evaluate the validity, the urinary flow patterns captured in these mobile-recorded micturition videos were compared with uroflowmetry metrics, including Qmax and voided volume. Observations were independently reviewed by two pediatric urologists, one adult urologist, and two residents to assess interobserver variability and ensure consistency across clinicians. Bland-Altman analysis was conducted to measure the agreement between video-based flow assessments and traditional machine-based uroflowmetry results.

The results highlighted strong interobserver reliability, with flow patterns such as Parabolic, Straight, and Waterfall achieving perfect agreement among clinicians (Intraclass Correlation Coefficient = 1.0). Bland-Altman analysis revealed minimal variability between observers, ensuring reproducibility. Additionally, video-based assessments showed strong correlations with traditional uroflowmetry metrics.

Parabolic flow patterns were associated with higher Q-max (peak flow rate), while Dribbling or Waterfall patterns indicated lower Q-max. Moderate correlations between Q-max and voided volume reflected individual variations in bladder capacity and voiding dynamics. Validation studies, such as Han et al. (2020), further confirmed the reliability of video-based recordings (rho = 0.778, p < 0.001), aligning with clinical expectations in pediatric and telemedicine settings.This study is limited by its focus on male pediatric patients, controlled clinical settings, reliance on clinician expertise, and smartphone video quality.

mobile-based video uroflowmetry can be a reliable, non-invasive alternative to conventional methods, demonstrating high interobserver agreement and strong correlations with standard metrics. Its adoption in low-income countries could address critical barriers such as cost, equipment availability, and healthcare access.