We are interested in the analysis of physiological and psychological impacts of vibration and acoustic noise on a person scanned in a weak-field magnetic resonance imager (MRI). The information about the state of the cardio-vascular system of a tested person is mainly acquired by current heart rate and arterial blood pressure (ABP) values. These parameters can be obtained by a photo-plethysmography (PPG) signal with parallel measurement by a blood pressure monitor (BPM). To enable proper and safe function in the weak magnetic field environment of the MRI device, special prototypes of wearable PPG sensors were developed. They consist of non-ferromagnetic materials and are fully shielded against the radiofrequency disturbance. If an external portable BPM device is applied for ABP measurement, the pressure cuff is always placed on the opposite arm than the PPG signal is sensed. From our latest analysis follows that also this measurement instrumentation arrangement has a partial influence on the properties of the sensed PPG wave.

Motivation of the current work was to find a different method to obtain ABP values excluding a BPM device to maintain the quality of the sensed PPG signal. In this paper we use an indirect approach where the ABP values are estimated from the sensed PPG wave. The proposed procedure uses time domain features (first of all the systolic/diastolic pulse times and their widths) extracted from the pre-processed PPG signal. Several different statistical approaches were tested to estimate the ABP value from these features. In the first step, the correctness of the estimated ABP values was confirmed by the BPM values measured in the frame of our previous measurements inside the running MRI device. Next, we plan to use the free access databases MIMIC or PPG-BP to verify the stability and accuracy of the developed ABP estimation method.