mardi 23 septembre 2025

par BUTEL Nathalie

Pascal Fechner, a former student of the UniCA M1 Computer Science program, is completing his thesis at the University of Bayreuth (Germany) and will be visiting the SIS center from September 15 to October 31. 

He will present his work and the instrument he has developed on Tuesday, September 23, 2025, at 10:30 a.m. at the i3s laboratory, room 007 of the Euclide B building.

 

Instrument 

An IoT-Enabled Wearable Device for Non-Invasive Urinary Bladder Monitoring Using Multi-Sensor Data and AI-Based Time Series Analysis

 

Problem

Neurogenic bladder dysfunction, frequently associated with spinal cord injury, multiple sclerosis, or spina bifida, severely impairs patients’ ability to perceive and regulate urinary bladder filling. This deficit often leads to medical complications and diminished quality of life, while current management strategies—such as intermittent catheterization or scheduled voiding—remain invasive, inconvenient, and insufficient to ensure autonomy.

 

Solution

Addressing this problem, an IoT-enabled wearable device for continuous, non-invasive bladder monitoring was developed by my research group at the University of Bayreuth (Germany). The system integrates multiple sensors, including near-infrared spectroscopy (NIRS), accelerometry, and temperature sensing, to capture physiological and contextual information. Edge computing capabilities enable local preprocessing and feature extraction, ensuring data privacy and reducing latency. Advanced AI methods, including Deep Learning architectures (LSTM, CNN), are employed for Time Series Analysis, enabling robust estimation and prediction of bladder filling levels. Transfer Learning strategies are further applied to account for inter-individual variability and improve model accuracy across different patient anatomies. The IoT-system communicates with a smartphone interface, allowing patients to access actionable feedback regarding bladder status and to manage voiding proactively.

In my presentation I will focus on both the IoT-related aspects and the signal processing of the device.