**5. Conclusions**

The results of the performed measurements help in precise description of the process of mechanical vibration excitation and the acoustic noise radiation in the scanning area and vicinity of the MRI device [22,23]. Comparison with a similar low-field MRI tomograph can be used for optimization of acoustic noise suppression in parallel with speech recording applied in 3D modeling of the human vocal tract [8]. The main usage of the obtained results is expected in experimental practice, when the used scan sequence and its parameters must often be changed depending on the person being tested. So far, our MRI experiments with vocal tract scanning were realized only on healthy people—typically, the trained MRI operators from the IMS and the authors of this paper themselves. In future, we plan cooperation with a medical center certified for work with patients. Then, our MRI equipment can be used for monitoring progress in treatment of diseases of the vocal cords and the vocal tract. For other users of this type of an open-air MRI device (in our country or in neighboring countries), we can only recommend choice of proper MR sequences, as well as modification of scan parameters as a compromise between noise and vibration exposition of an examined person and obtaining high-quality MR images.

We are aware of the limits of this study—at present, there exist a lot of modern MRI devices enabling parallel processing of MR images, thus rapidly shortening the necessary scanning time duration and subsequently minimizing the negative noise and vibration effect on patients. As the gradient coils are the integral part of the whole MRI equipment [6], we as users cannot make any changes to them; only external RF coils can be developed and connected to the system [24]. For the same reason, any system interference (e.g., adding new materials for damping of mechanical vibrations) cannot be recommended or advised as the output of this study.

Combination of the vibration and acoustic analysis can also be used in fault diagnosis of vibrating mechanical systems [25]. In our case, abnormal vibration and acoustic noise could reflect some fault in the MRI sequence generation. The performed measurement can also be reproducible on other types of low-field MRI devices; however, in the case of whole-body tomographs, picking up the vibration, as well as recording and measurement of the acoustic noise, are more difficult. In this case, the results obtained with an open-air device can be applied. A comparison study, including a discussion to this problem, has already been published in our last journal article [26].

The maximum noise SPL of about 78 dB (C) was measured at the closest distance from the central point of the MRI scanning area (45 cm), while the GE scan sequence with short TE and TR was running and the sagittal orientation was set. Although special hearing protection aids are not unconditionally necessary, ear plugs or ear muffs can contribute to the comfort of the examined person. If the scanned part of the human body inserted between the upper and the lower gradient coils is farther from the head, the ears are exposed by much lower noise. However, the absolute values of the vibration and noise signal RMS depend on signal amplification by the recording device, microphone directional pattern, and vibration sensor placement. The measurement of the noise itself is affected by a chosen measuring range (Lo/Hi) and a weighting filter (A/C) of the SPL meter. Therefore, direct comparison with the three relative energetic parameters calculated by Equations (3) and (4) is not possible. The scanning time duration depends on the chosen number of slices and their thickness. For 3D and Hi-Res sequences, it is usually lower than 15 minutes (typically about 3 to 5 minutes), so the vibration and noise exposition of the examined person does not mean a threat to his/her health. If more detailed MR images with higher quality factor *Q*<sup>F</sup> must be done (e.g., scans of particular parts of the human brain, the eye, the middle and inner ear, etc.), the patient will be exposed to the vibration and acoustic noise for a longer time *T*DUR (exceeding half an hour), which might be followed by a rather great physiological and psychological stress. Only urgent cases justify using these scan parameters in medical practice.

Our next research will be focused on detailed investigation of the negative influence of the generated acoustic noise on the physiological and psychological state of the examined person lying in the scanning area of the low-field MRI device. This negative influence on a human body can be monitored by measuring the blood pressure (BP) and heart rate (HR), as the stress is manifested by changes in the bloodstream. These changes can also be successfully determined directly from MR images of different parts of the human body—typically from the veins in the handbreadth area [27]. In addition, we plan parallel measurement of BP and HR parameters of the currently scanned person during execution of the whole MR scan sequence. For this purpose, we will apply photo-plethysmography (PPG) using optical sensors for non-invasive retrieval of vital information about the cardiovascular system from the skin surface [28,29]. Variations in the photo-detector signal are related to changes in the blood volume inside the tissue. Signal filtering and further processing will be necessary to obtain a clean PPG waveform, which can then be used to derive the instantaneous heart rate. This optical-based approach is fully in compliance with the requirements for sensors working in the magnetic field environment where RF and electromagnetic disturbances are present.

**Author Contributions:** Conception and design of the study, J.P., A.P., and I.F.; measurement, J.P.; data collection and processing, J.P.; writing, J.P. and A.P.); English correction, A.P.; paper review and advice, I.F.

**Funding:** This work was funded by the Slovak Scientific Grant Agency project VEGA 2/0001/17 and the Slovak Research and Development Agency, project no. APVV-15-0029.

**Acknowledgments:** We would like to thank all of our colleagues and other volunteers who participated in the listening test evaluation experiment.

**Conflicts of Interest:** The authors declare no conflicts of interest.
