Date of Award

August 2026

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Biomedical Engineering

First Advisor

Ryan Striker

Abstract

Foley catheter drainage systems remain largely passive, and assessment of urine passage, drainage interruption, and changes in urine appearance continues to depend on intermittent visual inspection, nursing documentation, caregiver observation, or delayed laboratory testing. Optical monitoring through the drainage-tube wall could provide objective information without opening the closed drainage pathway or placing sensors in direct contact with urine. This dissertation investigated the feasibility of a clamp-on multispectral optical platform for detecting urine passage and measuring wavelength-dependent optical changes under static and free-flowing drainage conditions.The investigation followed a staged medical-device development approach. Literature reviews characterized monitoring needs in acute, chronic, community, and long-term catheter care. A review of U.S. Food and Drug Administration post-market surveillance data further emphasized mechanical-integrity, flow and patency, leakage, usability, and sensing- and software-related failure modes existing in commercial Foley systems. These reviews inform the design and development of Foley system innovations, especially those related to measuring and monitoring urine quality. Monte Carlo light-transport model was adopted to simulate 365 nm light transmission through a multilayer optical path consisting of air, polyvinyl chloride tubing, urine, tubing, and air. The model demonstrated decreasing transmittance with increasing urine absorption and scattering and was compared with benchtop measurements of synthetic urine solutions. This establishes a repeatable method that can be used to test different optical paths and effects of light transmission through various mediums. A clamp-on, over-the-line prototype was developed using a Hamamatsu C12880MA microspectrometer, multispectral illumination near 388, 440, 500, 600, and 850 nm, Teensy 4.1-based acquisition, and a Python graphical user interface for visualization and data logging. Static performance was evaluated using pooled healthy human male urine prepared at seven water-dilution conditions with three independent runs per condition. The platform produced stable short-duration spectra through commercial drainage tubing, and the near-ultraviolet and violet spectral regions showed the greatest candidate sensitivity to dilution-related changes. However, wavelength ratios, principal component analysis, and cross-validated regression did not support reliable quantitative prediction of water dilution or urine specific gravity. Dynamic testing was performed using a benchtop bladder model, a Foley catheter, and five urine concentrations tested in triplicate. Multiband adaptive signal processing detected transient urine-passage events that were associated with visually observed drops. Principal component analysis showed differences between static and dynamic spectral measurements, although substantial overlap remained among urine concentrations. Run-level event features were insufficient to establish accurate urine-volume estimation. These findings demonstrate that non-contact multispectral sensing through Foley drainage tubing is feasible for stable short-duration spectral acquisition and dynamic urine-event detection. The current platform is not yet validated for urine-quality classification, specific-gravity prediction, constituent identification, or quantitative urine-output measurement. The work establishes a risk-informed and physics-informed engineering foundation for future calibration, long-duration verification, expanded biological testing, and clinical translation.

Download

Share

COinS