Real-Time Estimation of Bladder Residual Volume and Voiding Efficiency During Continuous Urodynamic Filling in Anesthetized Rats

Paquette, Jason 1 ; Yoo, Paul 1, 2  

1. Institute of Biomaterials & Biomedical Engineering, University of Toronto; 2. Edward S. Rogers Sr. Department of Electrical & Computer Engineering, University of Toronto

INTRODUCTION: Bladder residual volume (RV) and voiding efficiency (VE) are integral parameters used to monitor changes in urodynamic function in animal studies. Two complications that often arise during these studies stem from the unknown ureter urine contribution along with large errors introduced by repeated RV measurement after each void. It is important to accurately estimate these parameters to ensure reliable measurement of the urodynamic system. OBJECTIVES: The main objective of this study is to develop a simple approach for real-time estimation of the RV and VE of individual bladder voids during continuous-fill cystometry. This method would then be applied to in vivo experiments, in which changes in bladder function caused by tibial nerve stimulation (TNS) can be accurately monitored.

METHODS: Female Sprague-Dawley rats (250-300g) were initially induced with isoflurane (2-3%) for surgery. A PE-50 catheter was implanted and sutured into the bladder dome and a bipolar cuff electrode was implanted on the tibial nerve, located just above the ankle. Following all surgical procedures, the anesthesia was transitioned from isoflurane to urethane (1.2 g/kg, SQ). The rat was placed in prone position on our custom-fabricated experimental platform, which allowed real-time collection of the voided volume (VV). The bladder was initially emptied and then continuously filled at a constant rate (0.1 mL/min): 1-2 hour control period, 30 minutes of TNS, and 1-2 hour post-stimulation. At the end of each experiment, the bladder RV was obtained manually. The algorithm for estimating the RV began with an initial estimated value (neglecting ureters), which was subsequently refined by the following mass-balance equation: RV = RV(previous) +T*(PR+ UR) - VV; T: infusion time, PR: infusion pump rate, and UR: ureter fill rate. The RV estimates for each bladder void were continuously fitted against time using least squares regression (LSR). This enabled us to increase the accuracy of the estimated RV values by calculating UR estimates in an iterative manner. VE was calculated as follows: VE = VV/(RV+ VV).

RESULTS: Our real-time regression estimation method was tested in data obtained from 3 rat experiments, where the preliminary analysis was conducted on control period data. The fitted regression method estimated the UR in each animal: 3.5 uL/min (Rsq = 0.95), 6.1 uL/min (Rsq = 0.95), and 7.6 uL/min (Rsq = 0.97). The actual UR values in each experiment were obtained by directly calculating average UR from measured urodynamic variables. This yielded UR values of 3.4 uL/min (3% error from real-time estimate), 5.8 uL/min (5% error), and 7.3 uL/min (4% error), respectively.

CONCLUSIONS: Our preliminary analysis showed that our LSR method can estimate the UR in real-time with an error margin of 5%. Further analyses are to be performed on other urodynamic variables, both during and following TNS in these experiments.