Hypothesis / aims of study
We have developed a new method to validate and develop uroflowmetry devices. Uroflowmetry is a widely used urodynamic investigation tool in urology. The development of new urodynamic technologies requires a reliable, repeatable experimental environment to validate. The current standard for testing uroflowmeters is using a constant flow bottle to evaluate Qmax and Vvoid accuracy [1]. This is a simple and cheap method. However, it does not allow for testing the ability to accurately determine a given flow pattern, which is essential for designing novel techniques different from gravitational or rotating disk flowmeters. This study aims to develop and calibrate a low-cost laboratory setup capable of simulating uroflowmetry patterns.
Study design, materials and methods
We developed a pressure-driven flow setup utilizing a 150 ml pressure vessel to simulate the bladder. Pressure is regulated using a high-precision piezo-proportional pressure regulator (Festo VEAB, 0–1 bar range). Outflow is directed through 4 mm flexible tubing monitored by a clamp-on ultrasonic flow sensor (Keyence FD-Q). Closed-loop control was implemented using an Arduino microcontroller utilizing an open-source Proportional-Integral-Derivative (PID) algorithm (Arduino PID Library V1.2.1 by Brett Beauregard) [2].
The system was tested by tracking 5 curves representing a bell-shaped, tower-shaped, plateau, fluctuating and interrupted flow pattern according to ICS standards. The current setup simulates voiding patterns of young children (between 0-1 years of age) using parameters in ranges found by Olsen et al in 2009 [3].
Results
The tracked and reference flow curves before lag correction are shown in Figure 1. The mean results from the uroflowmetry simulations per flow pattern are shown in Figure 2. Showing reference peak flow (Qmax ref), measured peak flow (Qmax meas), reference volume (Vref), measured volume from the flow sensor (Vmeas), weighed volume (Vweighed), lag in seconds between measured and reference signal, weighted mean absolute percentage error (WMAPE) and the lag corrected WMAPE.
Interpretation of results
The system demonstrated high reproducibility across all five voiding patterns. Figure 2 shows small standard deviations in measured Qmax and derived volume. Noticeable is that the standard deviation is largest at the start of the measurement and at the end of the bell-shaped and tower-shaped pattern. Indicating the system requires some time to stabilize and has difficulty with very low flow rates (<1 ml/s). The difference between sensor-derived and gravimetrically weighed volumes remained below 2 mL across all patterns, confirming accurate flow sensor calibration.
Raw WMAPE values varied considerably between patterns (4.0–26.8%), which initially suggests inconsistent tracking performance. Cross-correlation analysis revealed a consistent temporal delay in the intermittent (0.32 s) and tower-shaped pattern (0.33 s), which have the highest peak flow rates and fast rising times. After lag correction, wMAPE values for all patterns fell below 16%, with the corrected values for Intermittent and tower-shaped dropping from 26.8% to 13.5% and 25.3% to 15.2% respectively.
For testing novel uroflowmetry technology, especially testing the reproducibility and accuracy are important. Both can be accurately tested using this setup.
The current setup is created for simulating pediatric flow patterns. The setup is scalable. To simulate voiding patterns of an older population the size of the vessel and outflow tube can be increased. This would result in a higher potential volume and Qmax.
This setup could also be used for different applications besides testing novel uroflowmetry solutions requiring simulation of uroflowmetry patterns.