Uroflowmetry derived from video urodynamic images automated using machine learning

van Dort W1, Rosier P1, Wortel R1, Schroeder R1, Geurts B2, de Kort L1

Research Type

Clinical

Abstract Category

Urodynamics

Best in Category Prize: Urodynamics
Abstract 145
Urodynamics
Scientific Podium Short Oral Session 17
Thursday 8th October 2026
14:30 - 14:37
Parallel Hall 4
Urodynamics Techniques Pediatrics Imaging Mathematical or statistical modelling
1. UMC Utrecht, 2. University of Twente
Links

Abstract

Hypothesis / aims of study
In patients who are unable to void on a uroflowmetry device, urethral resistance cannot be determined during urodynamic study. Videoflow is an algorithm that can derive uroflowmetry curves from standard care video urodynamic (VUDS) images [1]. In recently published versions of Videoflow the bladder was manually segmented on the VUDS images. The current study describes the validation of a fully automated machine learning (ML) process for segmentation and converting VUDS images into a uroflowmetry curve.
Study design, materials and methods
A multi-stage ML approach was chosen for segmentation of the bladder on VUDS images. Initially, a nnUNet model was trained on a random sample of 769 images taken from a dataset of 99218 VUDS images [2]. Subsequently, a second video object segmentation model (Cutie) was added to the analysis [3]. Cutie implements so-called memory frames, which are earlier segmentations of the object. This enables the ML algorithm to track an object over time, providing a starting point, which was given by the nnUNet segmentation. nnUNet and Cutie segmentation quality was assessed by the Dice Similarity Coefficient (DSC). Finally, the constructed Videoflow curve was compared with the standard uroflowmetry curve.
Results
Both segmentation stages showed an excellent performance with a median DSC of 0.965 after the first stage, significantly increasing to 0.972 after the second stage, see table 1. The greatest improvement was observed in the smaller bladder areas. The automated Videoflow showed similar to better performance compared to the manual Videoflow with similar maximum flow predictions, see figure 2.
Interpretation of results
The use of a multi-stage ML model for segmenting the bladder has a significant advantage over the single stage nnUNet. Especially for the smaller bladder areas, the sequential object tracking of Cutie resulted in a significantly better segmentation of the bladder. This better segmentation has a direct effect on the performance of the Videoflow algorithm, as visualized by the accuracy of the maximum flow. 

An interesting dependency of the segmentation accuracy on the bladder size was noted for automatic segmentation by the nnUNet algorithm. Especially for small bladder areas at the end of the voiding, the nnUNet did not recognize the bladder contour correctly, which resulted in huge flow spikes ‘constructed’ by the Videoflow algorithm directly on the nnUNet results. This dependency was also found in earlier studies implementing nnUNet, and is a generally acknowledged downside of UNet like ML algorithms. Due to the inherent down sampling process of the UNet architecture, small regions may be lost, resulting in poorer performance in smaller bladder areas. 

To overcome this, we used Cutie as a secondary segmentation stage, which implements object tracking over time. Use of Cutie resulted in a significantly better performance after the second stage, especially for the smaller bladder areas on the VUDS images. Although the combined ML model still showed  dependency on the segmentation accuracy of the segmented bladder area, this dependency was significantly reduced compared to using the nnUNet model alone.
Concluding message
In conclusion, we have validated the fully automated process of the Videoflow uroflowmetry curve prediction algorithm as a successor of the Videoflow algorithm that used manual segmentation. This novel algorithm, based on sequentially combined machine learning algorithms for bladder segmentation, showed similar good results as the manual method in generating uroflowmetry curves and no significant differences in Qmax compared to  standard uroflowmetry. This opens the possibility of applying the method in urological clinical practice, making it possible to generate uroflowmetry-curves in patients who are unable to void on standard uroflowmetry equipment. This technique allows urethral resistance to be determined in a group of patients for whom this was previously not possible.
Figure 1 Results after stage one and two of the median Dice similarity score and for each decile of segmented bladder area.
Figure 2 Visualization of the similarity of the standard uroflowmetry Qmax with the Videoflow Qmax, based on the manual segmentation, only nnUNet segmentation and the full segmentation pipeline including the Cutie segmentation.
References
  1. van Dort W, Rosier PFWM, Wortel RC, Schroeder RPJ, van Steenbergen TRF, Geurts BJ, de Kort LMO. Videoflow: Uroflowmetry in Children Exploiting Standard Care Video Urodynamic Imaging. Neurourol Urodyn. 2025 Nov;44(8):1569-1574. doi: 10.1002/nau.70135. Epub 2025 Aug 26
  2. Isensee, F., Jaeger, P.F., Kohl, S.A.A. et al. nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation. Nat Methods 18, 203–211 (2021).
  3. Cheng HK, Oh SW, Price B, Lee JY, Schwing A. Putting the Object Back into Video Object Segmenation. arXiv. 2024:2310.12982. doi: 10.48550/arXiv.2310.12982
Disclosures
Funding None Clinical Trial No Subjects Human Ethics not Req'd Retrospective study, checked by the local independent research office Helsinki Yes Informed Consent No AI Other AI Usage It is the core of the abstract. It was not used for textual purposes
06/06/2026 18:11:35