Near-Infrared Spectroscopy for Bladder Monitoring: A Machine Learning Approach

Dr. Jannik Lockl and Pascal Fechner published a scientific paper on the development and successful testing of an innovative alternative to conventional catheterization. The focus of this research was to develop a model for individualized bladder volume monitoring that could potentially improve the lives of millions of patients.

Patients living with neurogenic bladder dysfunction have no sense of their bladder or how full it is. To avoid over-distension and long-term urinary tract damage, the gold standard treatment is clean intermittent catheterization at predefined intervals. However, this schedule does not consider actual bladder volume, meaning catheterization is performed more often than necessary.

As part of the design science research process, the newly developed model was evaluated through interviews with affected patients, prototyping, and application to a real-world in vivo dataset. Using collected sensor data such as near-infrared spectroscopy and acceleration, the authors were able to predict bladder volume with an accuracy of 116.7 ml mean absolute error.

The promising results demonstrate the potential of the technology to improve bladder management, reduce invasive catheterization, and lower the use of medical aids and assistance from healthcare professionals.

The full article can be read here: https://dl.acm.org/doi/10.1145/3563779