Fiber-optic sensing of pulse wave velocity and magnetic field for synchronized cardiovascular monitoring in MRI

Abstract

Cardiovascular diseases (CVDs) represent a major global health challenge and are one of the leading causes of mortality worldwide. One of the key physiological parameters for CVD diagnostics is the Pulse Wave Velocity (PWV). This thesis focuses on the critical challenge of the ability to continuously record a patient's PWV under magnetic circumstances while enabling to synchronize with MRI by monitoring magnetic field change. The purpose is to design a wearable multi-position, highly accurate PWV monitoring system, which can operate and synchronize with MRI imaging from magnetic field detections.
To overcome the limitations of current PWV measuring devices, an invasive, multi-position Pulse Transit Time (PTT) measurement system using FBGs was developed. The system is integrated with Polydimethylsiloxane (PDMS) substrates and attached to six body locations, including left and right wrists, antecubital fossae, and feet. Data from 17 volunteers were collected and compared with a clinical reference device (Omron BP-203RPE III). To process low signal-to-noise ratio signals, five signal processing methods are evaluated, including systolic peak point-to-point, first and second derivative point-to-point, correlation coefficient, and a developed new phase difference Fourier transform (PDFT) method enhanced by an FFT resolution algorithm and five-point estimation. To synchronize with MRI data by environmental magnetic field detection, a magnetic field sensor based on magneto-shape effect of MF using a microfluidic chip and SPR fiber optics was developed and validated. The sensor detects meniscus position changes in a magnetic fluid-filled microchannel, altering the average refractive index sensed by firstly a TFBG, then a Surface Plasmon Resonance (SPR) sensor with Ag nano-arrays. The response to external refractive index changes and liquid surface movement was tested from experiments or FEM simulations. Because of the ineffectiveness of full-wave simulations in SPR optical fiber design, a computationally efficient semi-analytical Spatial Function Analysis Convolution (SAC) method was developed to model electromagnetic fields in nano-array fiber SPR sensors. By applying spatial convolution between single-element fields and distribution functions, the method reduces computational load significantly compared to full-wave FEM simulations, which was validated from generated Gaussian waveforms and FEM simulations.
The result shows that, for PWV system, with PDFT approach which achieved theoretical errors as low as ±1 ms, the lowest random error and failure rate (left: 7.63 ms, 12%; right: 15.07 ms, 10%) was observed among all of the methods. The system with PDFT method performed equivalently or better than the clinical device in baPWV evaluations. For magnetic field sensing system, sensor protype based on TFBG sensitivity reached 338.33 pm/RIU for refractive index and 1.18 pm/mm for displacement. By substituting with SPR sensors integrated nano-arrays, sensitivity was significantly enhanced to 2214.29 nm/RIU and 3.72 nm/μm for displacement, 6544 and 3.1×10⁵ times higher than TFBG, respectively. In terms of magnetic sensing, the system achieved 133 pm/Gs sensitivity (9.44×10⁻³ Gs⁻¹ intensity interrogation) with 6.7Gs error, 200 times more sensitive than conventional TFBG setups. For testing of SAC method, validation with generated signals showed residual variations below 2×10⁻⁶ V²/m² and errors under 0.5% in central regions. When applied to D-shape fiber nano-array sensor, the model achieved under 10% relative error with residual variance below 0.07 V²/m². Resource consumptions were significantly reduced, where storage requirements were presented as an example, reducing from an estimated 4 TB to 5 MB.
To conclude, an optical-fiber-based sensing system with capability of offering multi-position, accurate and robust monitoring of pulse wave velocity under magnetic environment, and has the ability to synchronize with data from MRI by monitoring environmental field change was established throughout the whole research work.

Date of Award	15 Mar 2026
Original language	English
Awarding Institution	University of Nottingham
Supervisor	Jing Wang (Supervisor), Richard Smith (Supervisor) & C.F. Kwong (Supervisor)