Lower urinary tract dysfunction (LUTD) refers to a range of disorders affecting the lower urinary tract and impacts millions of individuals worldwide, diminishing quality of life. Urodynamic studies (UDS), including cystometry, are used to diagnose LUTDs by measuring intravesical bladder pressure to assess bladder function. Conventional cystometry relies on a dual-catheter configuration, in which a second abdominal catheter serves as a reference signal to distinguish motion artifacts from true detrusor contractions. While effective, this approach increases invasiveness, is costly, and may limit bladder monitoring under natural conditions. Ambulatory cystometry enables real-world symptom replication during daily activities, but still relies on dual catheters.

This thesis investigates two strategies to enable motion context awareness in single-catheter ambulatory cystometry, aiming to reduce invasiveness, improve accessibility, and maintain diagnostic reliability.

Part A explores a machine learning–based approach to distinguish motion artifacts from detrusor contractions using a single pressure signal. A structured, posture-aware signal processing and classification pipeline was developed and evaluated using cystometry data from ten patients. Multiple modelling strategies were evaluated, including feature-based classifiers, neural network architectures, and hybrid strategies. Neural networks demonstrated superior overall performance, although increasing architectural complexity did not result in noticeable performance gains. While results are limited by cohort size, the findings indicate that motion artifacts and detrusor events exhibit distinguishable pressure characteristics and establish a scalable analytical framework for future dataset expansion.

Part B addresses motion context awareness through hardware integration. The UroMonitor, a low-cost, single-catheter ambulatory cystometry prototype, was independently designed and developed. The system integrates a pressure sensor, inertial measurement unit (IMU), microcontroller, wireless connectivity, and on-board storage within a portable, battery-powered platform. By simultaneously acquiring pressure and motion signals, the device enables correlation-based identification and suppression of motion artifacts without requiring a second catheter. Phantom testing demonstrated reliable motion detection, with acceleration magnitude signals achieving a mean recall of 95.4\%. The prototype satisfied predefined physical, electrical, economic, physiological, and usability design constraints.

Together, these contributions demonstrate the technical feasibility of a low-cost, motion-aware, single-catheter ambulatory cystometry system through both machine learning and hardware-based strategies, providing a foundation for future clinical validation and scalable deployment, including in low-resource healthcare settings.