Our research is oriented to analysis how exposition of mechanical vibration and acoustic noise generated during scanning by a magnetic resonance imaging (MRI) tomograph impacts cardiovascular system of examined persons. This negative effect is expressed by changes in arterial stiffness, heart rate (HR), arterial blood pressure, and/or other parameters. Optical sensors working on a photoplethysmography (PPG) principle are well applicable to monitor these changes.
Current work focuses on investigation of basic characteristics, testing, and comparison of three wearable multi-sensors with Bluetooth (BT) data transfer. The developed special prototypes based on optical PPG sensors contain also physical sensors of temperature and humidity to monitor actual state of skin, and a force-sensitive-resistor (FSR) component primarily used for proper contact pressure adjustment, applicable also to sense a heart pulsation. All multi-sensor parts consist of non-ferromagnetic elements and are fully shielded to be practically usable in planned measurements in the low magnetic field environment with additional radio frequency disturbance.
The paper describes three sets of experiments: (i) Precision and short/long-time stability of physical sensors using certified measuring devices were analyzed in air-conditioned temperature and humidity environments. (ii) BT connectivity and data transfer to master control device was tested inside the working MRI tomograph without any human subject. (iii) Comparative measurements with a clinical pulse oximeter on different fingers of testing persons in normal laboratory conditions were aimed to evaluate the precision of HR values determined from PPG (HRPPG) and FSR (HRFSR) signals. Statistical parameters of HRPPG sequences reach better properties than HRFSR ones. Essential differences of mean absolute error (MAE) are found for PPG signals sensed on left and right hands, while the Pearson’s correlation coefficient Rcc differs minimally. Summary results show acceptable accuracy with MAE below 2 bpm and mean Rcc about 0.97, which is principally comparable with other PPG-based measuring systems.