Using Real-Time Artificial Intelligence to Improve Bladder Tumor Detection and Characterization in Patients Undergoing Cystoscopy

What this Trial Is About

Researchers are evaluating a new method to improve the diagnosis and monitoring of bladder tumors by combining traditional bladder endoscopy with an artificial intelligence AI based diagnostic tool called the Bladder-Portable Artifact Detection System Bladder-PAD. The study aims to reduce diagnostic errors and standardize tumor descriptions to better characterize tumor aggressiveness and improve patient care. This observational study is led by the Centre Hospitalier Universitaire, Amiens. The study uses video frames extracted from a dataset of cystoscopy videos, which are then analyzed using a pre-trained convolutional neural network model with feature detection algorithms. The Bladder-PAD system will be assessed for its ability to detect bladder tumors and predict the risk of tumor recurrence and progression compared to standard white light cystoscopy. Participants will undergo cystoscopy examinations where tumor detection rates and false detection rates will be measured for both white light cystoscopy and Bladder-PAD cystoscopy on the same day. Data analysis will occur at both patient and lesion levels to evaluate the accuracy of the AI system. The study involves no treatment intervention but focuses on diagnostic performance, aiming to develop a precise diagnostic companion tool for bladder tumor management.

Augmented Bladder Tumor Detection Using Real Time Based Artificial Intelligence