-
- Vrsta/tip rada: Pregledni (revijski) članak
- PREGLEDAJTE RAD
- PREUZMITE RAD
-
- 76 pregleda
- 585 puta preuzet
Abstract
Introduction. Urodynamic studies are a fundamental diagnostic tool in functional urology, crucial for assessing urinary tract function. Over the past decade, artificial intelligence has been progressively applied across several scientific disciplines, with healthcare representing a significant area of growth. This review aimed to examine the published literature on the use of machine learning in urodynamics over the preceding five years.
Material and Methods. A comprehensive literature review was conducted using Google Scholar and PubMed to identify studies published between 2020 and 2025 on AI applications aimed at reducing the invasiveness of urodynamic testing, enhancing diagnostics, and predicting treatment outcomes.
Results. Recent studies demonstrate that artificial intelligence and machine learning significantly enhance diagnostic precision and predictive capability in urodynamics. Algorithms including convolutional neural net works, support vector machines, logistic regression, and decision trees accurately classify detrusor overactivity, bladder outlet obstruction, and other urodynamic abnormalities.
Discussion. Deep learning models excel in analysing complex high-dimensional data, achieving reporting accuracies of up to 100%, whilst conventional machine learning methods provide greater interpretability. AI systems using uroflowmetry data reached diagnostic accuracies of up to 84%, surpassing expert performance.
Conclusion. Overall, artificial intelligence and machine learning improve the reliability and efficiency of urodynamic assessments, although larger clinical studies are required for validation.