Open Access
Add to BibSonomy
Abstract
Continuous noninvasive measurement of vital bio-signs, such as cardiovascular parameters, is an important tool in evaluation of the patient’s physiological condition and health monitoring. Based on new enabling technologies, continuous monitoring of heart and respiration rate, pulse wave velocity and blood pressure have been investigated, advanced and reflected in numerous papers published in recent years. In this paper, we introduce a new technique for noninvasive sensing of vital bio-signs based on a multimode optical fiber sensor that can be integrated into a fabric. The sensor consists of a laser, optical fiber, video camera and computer. Its operation is based on tracking of point-wise intensity variations on speckle patterns caused by interference of the light modes within the fiber subjected to deformation. The paper contains theoretical analysis and experimental validation of the proposed scheme. The main goal is to advance a simple low-cost sensor embedded in a cloth fabric to track changes in the cardiovascular condition of the wearer.
© 2020 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Continuous measurement of cardiopulmonary parameters, such as respiratory rate (RR) and heart rate (HR), is crucial during anesthesia, for diagnosing heart diseases such as stroke, heart failure, heart attack and hypertension, as well as anomalies such as hyperventilation and sleep apnea [1–7]. It is also important for daily healthcare monitoring, especially in infants, the elderly, people with special needs, and patients requiring continuous medical assistance and treatment.
Application of Optical Fiber Sensors (OFS) is an attractive way to measure these vital bio-signs. The high information transmission capacity of optical fibers, combined with additional advantages, such as intrinsically safe modes of operation, inertness to chemicals, nontoxic origin and light weight, are the main factors attracting the attention of researchers [8]. Optical fibers also have well-known immunity to electromagnetic radiation, which makes them suitable for healthcare monitoring in an MRI environment. Since the sensors are in a diagnostic environment and in proximity to biomedical devices, it is important that they not be susceptible to electrical discharges or generate heat.
OFS can be broadly classified as extrinsic or intrinsic. Extrinsic OFS carry light to an external sensing system, while intrinsic OFS perform the sensing operation inside the fiber. OFS may be grouped into four main categories, depending on the light characteristics being modified [9]. These categories include sensors based on intensity modulation, phase modulation (interferometry), polarization modulation (polarimetry) and wavelength modulation (spectrometry). Respiration and heartbeat monitoring by OFS can be done using a variety of instruments and methods [10–17]. Several research groups have investigated Bragg grating wavelength modulated sensors for monitoring respiratory and cardiac activity [12,13,18,19]. Nevertheless, wavelength detection–based technology is expensive and too complicated for both fabrication and implementation.
The Mach-Zehnder interferometer (MZI)–based sensor belongs to the phase modulated category, where a single mode in-fiber MZI is used to measure RR, HR and pulse wave velocity (PWV) [20]. However, due to the singularity of the light mode, it is necessary to use both a control arm and a sensing arm to produce interference of light in the detector. Another way to measure vital bio-signs is by using multimode fibers (MMF). Due to the several modes inherent in the MMF, fiber deformation caused by body vibration changes the mode interaction and coherent light intensity (intensity modulated sensor).
Detection of RR and HR using MMF has been the subject of numerous papers referring to aspects such as the monitoring of heartbeats and respiration using an optical fiber micro bend sensor embedded in textiles, pillows or bed mattresses [7,21–24]. The working principle behind this sensor is based on the theory of optical fiber micro bending. Although separate sensing of respiration rate and heart rate has been previously demonstrated, other parameters such as PWV have not been measured using MMF sensors. The previous works show that MMF are affected by external noise, and further algorithm development and signal filtering are recommended to improve sensor performance.
The configuration of the approach presented herein to monitor vital bio-signs is based on a laser, MMF, single video camera and computer. According to the proposed technique RR, HR and PWV can be detected simultaneously and non-invasively. MMF with a step index was chosen to generate speckle patterns in the fiber output due to interference between the different modes within the fiber subjected to deformation [25]. Such configuration eliminates the need for an additional reference arm when using a single-mode fiber sensor. The MMF sensor introduced in our work is a single arm device. Under deformation of the fiber due to heart palpitations / respiratory expansions, the coherent light within the MMF is modulated, the light modes interfere, and the output speckle pattern varies in its point-wise intensity. A defocused camera [26,27] captures these changes. Due to defocusing (performs an approximated Fourier transform) the variation of the speckle pattern at the output from the MMF is better observed as linear phase change which after defocusing is converted to shifts in the camera plane. Indeed, the speckle pattern do not have only to shift, it can also change. However, some energetic portion of it will shift. Any linear phase generation component obtained due to some perturbations in the fiber, will lead to shift in the camera plane since the optics of the camera was defocused and thus it performs a Fourier transform over the light distribution obtained at the output of the fiber. Processing the speckle pattern images allows extraction of the RR, HR and PWV.
2. Theoretical background
2.1. Multimode fiber speckle pattern
When coherent light propagates through an MMF, the created multiple modes, considered to be equally excited, interfere and create a speckle pattern. The fiber far-field light speckle distribution $({A_0})$ is the superposition of all the modes’ amplitudes, as given by [28]:
The far-field speckle pattern intensity $I({x,y} ),$ captured by a defocused camera [26] from the edge of the fiber, may be described as [28]:
Deformation of the fiber causes perturbation in the propagation medium of each mode. The modes are affected by the perturbation differently: amplitude and phase deviate depending on the mode index. The speckle pattern image changes with fiber deformation.
When the fiber is perturbed by some physical means, such as respiration / heart rate, tracking and processing of speckle pattern intensity can allow vital bio-signs to be determined and recorded.
The average of point-wise intensity changes observed between two adjacent frames may be presented as:
The intensity should change since the processing takes place in the far field of the speckle patterns, thus change in linear phase causes movement of the speckle pattern obtained in the camera plane and therefore in the camera plane there is a change in the intensity of the difference between spatial distributions in sequential frames.
2.2. Evaluation of pulse wave velocity and pulse pressure by multimode fiber sensor
Due to the periodic nature of the cardiac cycle, causing transitional blood flow, the pulse wave propagates through the arterial tree from the heart to the periphery. Pulse wave velocity is widely recognized as one of the vital cardiovascular bio-signs. Elevated PWV is commonly acknowledged to be a bio marker of atherosclerosis, hypertension and coronary heart disease [29–34]. PWV may be evaluated by recording the Pressure-wave Transition Time (PTT) between the two selected areas in the arterial tree, e.g. heart and wrist.
The average PWV may be estimated as [35]:
where L is the arterial length between heart and wrist.According to the Moens-Korteweg equation [36]:
where E is the Young modulus of the arterial wall, d is the blood vessel diameter, h is the wall thickness, and ρ is the blood density.An increase in arterial stiffness and loss of elasticity with age will elevate the PWV in accordance with Eq. (5). The threshold value for PWV is around 10 m/s. The PWV varies not only during the day, but also within the cardiac cycle. An increase in blood pressure causes temporary stiffening of the vessels, which can result in a higher PWV. The vessel diameter also varies during the cardiac cycle, depending on the vessel's elasticity, and is related to blood pressure. In light of the above, PWV is widely used to estimate pulse pressure (PP) and systolic blood pressure (SP) [37,38].
According to Zukovsky, pulse pressure (surge pressure) may be presented as [39]:
After rearranging the Bramwell and Hill equation [40,41]:
where Δp is pulse pressure, Vmax is the aortal maximum blood flow velocity, d is the blood vessel diameter, and ρ is blood density.Equations (6) and (7) show that PWV may be used as a major indicator of blood pressure.
3. Materials and methods
3.1. Experimental setup
3.1.1. Pulse wave and heart-beat detection by multimode fiber sensor
The setup of the experiment is presented in Fig. 1.
Fig. 1. Pulse wave detection configuration. a) Wrist sensing. b) Laser and the illuminated fiber. c) The heart and wrist simultaneously detected speckle patterns. d) Heart sensing. e) The fibers pointed towards the camera in order to capture the speckle patterns.
The system consists of II-VI Suwtech DPGL-2100 532 nm green laser, acA1920-25um Basler camera with focal length of 8 mm and F# of 1.4 and FP400ERT non-tubing 0.50 NA, Ø400 µm core multimode fiber with spectral range of 400-2400 nm. The laser is positioned 18 cm (but could be varied) from the end of the illuminated fiber. The sensing part of the MMF is fastened with a strap to the area of the heart in the subject’s chest. The sensing part of the fiber is configured in four loops in order to improve sensitivity. The other end of the fiber is pointed towards the camera in order to capture the output light forming the speckle patterns.
The tested subject was requested to hold breathing for the duration of the recording, which lasted 10 seconds. The speckle patterns were detected by the camera working at a frequency of 100 frames per second (FPS). The results were analyzed using signal processing algorithms in order to evaluate the change in the point-wise intensity of the speckle patterns over time. The video recordings were analyzed using image processing algorithms. Several algorithms were developed, verified and compared in order to determine the most effective algorithm for tracking changes in the speckle patterns (see paragraph 4.1.2 [26,28,35]).
3.1.2. Respiratory rate detection
The same system configuration, with the sensing part of the fiber attached to the subject’s chest, was used for respiration rate detection. For the 10 seconds recording time the subject was requested to stand still and breathe normally. As the subject exhaled and inhaled, the thoracic pressure changed, thus applying pressure on the attached fiber sensor, causing its perturbation and breath detection.
3.1.3. Simultaneous detection of respiratory and heart rates
The sensing part of the fiber was attached to the subject’s chest. For the 10 seconds recording time the subject was requested to breathe normally. Two different perturbations were simultaneously detected by the same fiber: MMF perturbation caused by heart contraction / relaxation, and RR thoracic pressure variation.
3.1.4. Pulse wave velocity estimation
The structure of the system is illustrated in Fig. 2. Pulse wave velocity was estimated based on the simultaneous pulse wave recording from the heart and wrist areas using two similar MMF placed accordingly.
Fig. 2. Implemented optical configuration for remote measurement of heart rate, respiration rate and pulse wave velocity. a) Ref. [42] b) Ref. [43] c) Ref. [44].
In order to achieve time synchronization, the 50:50 1X2 fiber coupler was used to split the light that was injected from the laser to the two identical MMF. The first fiber is attached to the front chest near the subject's heart area. The second fiber is attached to the subject's wrist, near the radial artery. To avoid the need to synchronize two cameras, as depicted in the previous work [20], the light radiated from both fibers is captured by a single camera, so that the speckle patterns from the two fibers are projected without overlapping [Fig. 1(C)].
The tested subject was requested to hold breathing for the duration of the measurement. The same algorithm was applied for processing the two speckle patterns. The PWV and the PTT were calculated by subtracting the time between the peaks detected from the two fiber sensors. The length of the arterial tree between heart and wrist was determined by approximating arterial tree and measuring it with flexible ruler for each person.
In order to determine the relationship between PWV and SP, or PWV and PP (the difference between systolic and diastolic blood pressure), blood pressure was measured for each video recording using a commercial Blood Pressure (BP) monitor (Omron MX2 basic). The process included measuring the PWV while the subject's BP varied after exercise from elevated to steady level.
4. Results
4.1. Optimization of MMF sensor performance
4.1.1. Influence of fiber type on sensor performance
Operation of the MMF sensor depends on parameters such as fiber core diameter, coating and Numerical Aperture (NA). To determine the optimal fiber type, a number of MMF were selected and evaluated (see Table 1). All selected multimode fibers had NA = 0.5.
Table 1. Summary of the tested multimode fibers.
Two sensors having different fibers were attached to the chest near the heart area. The speckle patterns produced by the sensors were simultaneously recorded and processed with the selected algorithm. The results show that the Ø400 µm fiber without coating produced the best combination of sensitivity and SNR.
4.1.2 Selection and development of the speckle pattern–processing algorithm
To determine the optimal signal-processing algorithm we applied the following procedures:
- 1. Sum of Absolute Difference (SAD) of the speckle pattern pixels between consecutive frame images. The pixel intensity of the first out of two adjacent images is subtracted from the corresponding pixel intensity of the other image and the absolute values are totaled.
- 2. Correlation of the speckle pattern images in order to determine the displacement between the images.
- 3. Sum of the Absolute Value (SAV) of the first-time derivative of the speckle pattern pixels gray intensity between consecutive frame images.
- 4. SAV of binary speckle pattern images. Binarization is applied by the authors in order to eliminate the effect of speckle shape variations within an image. The proposed binarization algorithm allows the speckle locations to be determined and the representative image of the corresponding binary points to be built. Image binarization is followed by correlation / SAV speckle pattern processing.
- 5. Application of the Kalman filter after SAV image processing in order to reduce the effect of noise.
- 6. Application of a standard optical flow algorithm based on the distribution of apparent velocities of the speckle patterns.
The SAV algorithm [see Eq. (4)] was selected based on a comparison of the processed signals – this being the fastest and showing the highest SNR. It should be mentioned that applying the Kalman and the common filters caused a non-regular shift in corresponding wave peaks, affecting the precision of PWV evaluation.
4.2. Results of cardiovascular bio-signs detection
4.2.1. Heartbeat detection by MMF sensor
The results of pulse wave detection by the MMF sensor (heart area) are presented in Fig. 4. Each heartbeat cycle is determined by clearly detectable and repeatable waveforms containing two dominating peaks related to the first and second heart sounds. It was assumed that the peaks are related to S1 and S2 due to the fact that they are repeatable having corresponding time delay between the events. S1 is caused by closure of the mitral and tricuspid valves at the beginning of ventricular contraction; S2 is caused by closure of the aortic and pulmonic valves at the beginning of ventricular relaxation [45]. The difference between S1 and S2 was 0.2 sec.
Coefficients of variation of ΔS1, ΔS2 and S2-S1 are within 5% range. The heartbeat rate was calculated based on the time difference between the corresponding wave-peaks related to the maximum speckle pattern point-wise intensity variation. An example of intensity variation during the heartbeat detection experiment is presented in Fig. 5. The results of the heart rate detection were compared with direct measurement using palpation and a commercial blood pressure monitor (Omron MX2 basic). The average of the heartbeat rates, standard deviation and Coefficient of Variation (CV) were calculated. The duration of the calculated average heartbeat cycle was 0.88 ± 0.02 sec with CV = 0.02/0.88 = 0.023, corresponding to an HR of ∼68.18. The heart rate measured by the blood pressure monitor was 70.
Fig. 5. Speckle pattern intensity variation during the heartbeat detection experiment. Frame (a) was in the lower intensity range compared with frame (b) showing one of the peak intensities (The same measurement as per Fig. 4).
4.2.2. Pulse wave detection from the wrist area
Pulse wave detection from the wrist area is presented in Fig. 6.
A recording of the pulse wave from the wrist area shows a single peak prevailing during each heartbeat cycle. The regular difference between the waveforms was also observed. The heartbeat rate was calculated based on the time difference between the corresponding wave peaks. The results were verified using a certified medical reference device. The heartbeat mean value, STD and CV were calculated. The average time difference between the peaks was 0.96 ± 0.06 sec, CV = 0.063, with a corresponding HR of ∼62. The direct measured HR of the subject was 60.
4.2.3. Breath detection
The MMF sensor was attached to the chest near the heart area.
Fast breathing cycles (inhalation and exhalation) were recorded along a 30-second period. During the experiment the breathing cycle started from inhalation.
We recorded start of inhalation and found that it corresponds to the visible sharp raise of amplitudes in the signal graph presented in Fig. 6. The recorded period of breathing and its number also correspond to this raise as it is shown on the graph. The results presented in Fig. 7, showing waveforms related to the respiratory cycles. The respiratory rate duration and time between breathing cycles can be extracted from the recordings. The respiratory rate for the tested subject was 12 breaths per minute.
Fig. 7. Fast breathing detection with sampling frequency of 100 Hz with the sensing part connected to the heart area.
4.2.4. Simultaneous detection of heart and breath rate
The tested subject was requested to breathe normally during a 30-second period. The test results are presented in Fig. 8, showing that the breathing rate and heart rate may be detected simultaneously.
The frequency domain is presented in Fig. 9, showing that the heart rate frequency is ∼1 Hz and the respiratory rate frequency ∼0.2 Hz, corresponding to a heart rate of 60 beats/min and respiratory rate of 12 breathings/min.
4.2.5. Pulse wave velocity and blood pressure estimation
In order to determine the PWV, pulse waveforms from the heart and wrist areas were recorded simultaneously (see Fig. 10). The time delay between the waveforms is related to the pressure wave propagation between two measurement sites (heart and wrist). The wave peak from the heart area (red) precedes the wave peak from the wrist area (blue).
The time delay between the peaks allows calculation of the PTT. The PWV can be calculated taking into consideration the length of the arterial tree pass between heart and wrist.
The average value of the PTT obtained was 0.075 sec. The estimated length of the arterial segment was 0.75 m. The PWV was estimated as 10 ± 0.04 m/sec, CV = 0.004.
The PWV measured by the two-sensor system allows monitoring of the condition of the arterial tree. Embedded in a cloth fabric, the sensors allow continuous monitoring of the cardiovascular system. Due to the strong correlation between the PWV and blood pressure [46,47], PWV monitoring enabled evaluation of blood pressure.
In order to determine the relationship between BP and PWV by means of the dual MMF system described, five individuals aged 26-73 were tested. The results were approximated and validated by linear regression models are they presented in Figs. 11 and 12. Correlation between the PP, BP and PWV is found in the range of values of 0.84-0.99. Thus, linear regression model gives a good approximation to the experimentally measured results.
The recordings were conducted after exercise to raise the blood pressure of the individuals. Under exercise the cardiac output and blood velocities are both increasing, and this is causing additional flow resistance and consequent raise of blood pressure.
The BP was measured in parallel with the pulse waves recording. The measurements were repeated several times up to the point where the BP returned to the pre-test values. The relationship between the PWV and pulse pressure, and the PWV and systolic pressure for the five tested individuals is presented in Fig. 11 and Fig. 12, respectively.
The PWV uncertainty can be observed from the red colored readings (male 73). For the other tested subjects, the uncertainty was negligible. Deviation of the obtained results from the linear model could be related to the arterial diameter variation which is not included in the regression model but exists in Eq. (6). Linear approximation of the relation between systolic blood pressure and PWV for the 5 tested individuals shows high level of correlation.
It may be seen that the same trend exists for all tested subjects: a rise in PP results in a higher PWV. Pulse pressure and PWV have a tendency to increase with age, which is in full compliance with previous works [41,48,49].
5. Conclusions
The present work contains further development of the MMF sensor for biomedical applications. The sensor consists of a laser, MMF, a single digital camera and a computer. The sensor is based on tracking the point-wise intensity of the speckle patterns radiating from deformed MMF and captured by digital camera. The novelty of the sensor part is based on the following:
- • A 400 µm diameter fiber without coating was found to have the best combination of sensitivity and SNR.
- • Optical synchronization of the two speckle patterns and related signals was achieved by using a single digital camera.
- • The authors evaluated and compared speckle patterns’ processing algorithms and found that the SAV of the first-time derivative algorithm provides the highest sensitivity and SNR. The developed algorithm allows simultaneous detection of heartbeats and respiration rate with high precision. It allows to evaluate heart beat variation related to arrhythmia. The average heart rate variation of 2.5% that is presented in this paper could be related to the level of precision of the validation instrument as well as to the variation of the cardiac cycle duration.
- • The authors introduced for the first time an MMF sensor for evaluation of PWV, commonly recognized as a prognostic bio-marker of arteriosclerosis. Monitoring of PWV also allows estimation of the systolic blood pressure. The MMF embedded in fabric could lead to the development of “smart clothes” for continuous monitoring of vital bio-signs, such as heart and respiratory rate and blood pressure, as part of home and ambulatory health care improvement.
Disclosures
The authors declare no conflicts of interest.
References
1. J. F. Fieselmann, M. S. Hendryx, C. M. Helms, and D. S. Wakefield, “Respiratory rate predicts cardiopulmonary arrest for internal medicine inpatients,” J. Gen. Intern. Med. 8(7), 354–360 (1993). [CrossRef]
2. O. S. Hoilett, A. M. Twibell, R. Srivastava, and J. C. Linnes, “Kick LL: A Smartwatch for Monitoring Respiration and Heart Rate using Photoplethysmography,” in 2018 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), pp. 3821–3824 (2018).
3. L. J. Epstein, D. Kristo, P. J. Strollo, N. Friedman, A. Malhotra, S. P. Patil, K. Ramar, R. Rogers, R. J. Schwab, and E. M. Weaver, “Clinical guideline for the evaluation, management and long-term care of obstructive sleep apnea in adults,” J. Clin. Sleep Med. 05(03), 263–276 (2009). [CrossRef]
4. H. Folgering, “The pathophysiology of hyperventilation syndrome,” Monaldi archives for chest disease = Archivio Monaldi per le malattie del torace 54, 365–372 (1999).
5. J. C. Missri and S. Alexander, “Hyperventilation syndrome: A brief review,” JAMA 240(19), 2093–2096 (1978). [CrossRef]
6. D. S. Cowley and P. P. Roy-Byrne, “Hyperventilation and panic disorder,” Am. J. Med. 83(5), 929–937 (1987). [CrossRef]
7. X. Yang, Z. Chen, C. S. M. Elvin, L. H. Y. Janice, S. H. Ng, J. T. Teo, and R. Wu, “Textile fiber optic microbend sensor used for heartbeat and respiration monitoring,” IEEE Sens. J. 15(2), 757–761 (2015). [CrossRef]
8. M. Szustakowski and N. Palka, “Contrast sensitive fiber optic Michelson interferometer as elongation sensor,” Opto-Electron. Rev. 13, 19–26 (2005).
9. G. Rajan, Optical fiber sensors: advanced techniques and applications, CRC press (2015).
10. A. Grillet, D. Kinet, J. Witt, M. Schukar, K. Krebber, F. Pirotte, and A. Depré, “Optical fiber sensors embedded into medical textiles for healthcare monitoring,” IEEE Sens. J. 8(7), 1215–1222 (2008). [CrossRef]
11. M. Nishyama, M. Miyamoto, and K. Watanabe, “Respiration and body movement analysis during sleep in bed using hetero-core fiber optic pressure sensors without constraint to human activity,” J. Biomed. Opt. 16(1), 017002 (2011). [CrossRef]
12. L. Dziuda, F. W. Skibniewski, M. Krej, and J. Lewandowski, “Monitoring respiration and cardiac activity using fiber Bragg grating-based sensor,” IEEE Trans. Biomed. Eng. 59(7), 1934–1942 (2012). [CrossRef]
13. L. Dziuda, F. W. Skibniewski, M. Krej, and P. M. Baran, “Fiber Bragg grating-based sensor for monitoring respiration and heart activity during magnetic resonance imaging examinations,” J. Biomed. Opt. 18(5), 057006 (2013). [CrossRef]
14. W. SpillmanJr, M. Mayer, J. Bennett, J. Gong, K. Meissner, B. Davis, R. Claus, A. MuelenaerJr, and X. Xu, “A ‘smart’bed for non-intrusive monitoring of patient physiological factors,” Meas. Sci. Technol. 15(8), 1614–1620 (2004). [CrossRef]
15. T. D. Allsop, T. Earthrowl, R. Revees, D. J. Webb, M. Miller, B. W. Jones, and I. Bennion, “Application of long-period grating sensors to respiratory function monitoring,” in Smart Medical and Biomedical Sensor Technology II (International Society for Optics and Photonics), pp. 148–156 (2004).
16. J. Wo, H. Wang, Q. Sun, P. P. Shum, and D. Liu, “Noninvasive respiration movement sensor based on distributed Bragg reflector fiber laser with beat frequency interrogation,” J. Biomed. Opt. 19(1), 017003 (2014). [CrossRef]
17. S. Šprager and D. Zazula, “Detection of heartbeat and respiration from optical interferometric signal by using wavelet transform,” Comput. Meth. Prog. Bio. 111(1), 41–51 (2013). [CrossRef]
18. M. Ciocchetti, C. Massaroni, P. Saccomandi, M. Caponero, A. Polimadei, D. Formica, and E. Schena, “Smart textile based on fiber bragg grating sensors for respiratory monitoring: Design and preliminary trials,” Biosensors 5(3), 602–615 (2015). [CrossRef]
19. C. Massaroni, P. Saccomandi, D. Formica, D. L. Presti, M. A. Caponero, G. Di Tomaso, F. Giurazza, M. Muto, and E. Schena, “Design and feasibility assessment of a magnetic resonance-compatible smart textile based on fiber Bragg grating sensors for respiratory monitoring,” IEEE Sens. J. 16(22), 8103–8110 (2016). [CrossRef]
20. T. Sirkis, Y. Beiderman, S. Agdarov, Y. Beiderman, and Z. Zalevsky, “Fiber sensor for non-contact estimation of vital bio-signs,” Opt. Commun. 391, 63–67 (2017). [CrossRef]
21. Z. Chen, J. T. Teo, and X. Yang, “In-bed fibre optic breathing and movement sensor for non-intrusive monitoring,” in Optical Fibers and Sensors for Medical Diagnostics and Treatment Applications IX (International Society for Optics and Photonics), p. 71730P (2009).
22. Z. Chen, J. T. Teo, S. H. Ng, and X. Yang, “Plastic optical fiber microbend sensor used as breathing sensor,” in SENSORS, 2012 IEEE, pp. 1–4 (2012).
23. C. J. Deepu, Z. Chen, J. T. Teo, S. H. Ng, X. Yang, and Y. Lian, “A smart cushion for real-time heart rate monitoring,” in 2012 IEEE biomedical circuits and systems conference (BioCAS), pp. 53–56 (2012).
24. Z. Chen, D. Lau, J. T. Teo, S. H. Ng, X. Yang, and P. L. Kei, “Simultaneous measurement of breathing rate and heart rate using a microbend multimode fiber optic sensor,” J. Biomed. Opt. 19(5), 057001 (2014). [CrossRef]
25. J. W. Berthold, “Historical review of microbend fiber-optic sensors,” J. Lightwave Technol. 13(7), 1193–1199 (1995). [CrossRef]
26. Z. Zalevsky, Y. Beiderman, I. Margalit, S. Gingold, M. Teicher, V. Mico, and J. Garcia, “Simultaneous remote extraction of multiple speech sources and heart beats from secondary speckles pattern,” Opt. Express 17(24), 21566–21580 (2009). [CrossRef]
27. Z. Zalevsky and J. Garcia, “Motion detection system and method,” US patent No. 8638991 (2014).
28. T. Francis, M. Wen, S. Yin, and C. M. Uang, “Submicrometer displacement sensing using inner-product multimode fiber speckle fields,” Appl. Opt. 32(25), 4685–4689 (1993). [CrossRef]
29. H. L. Kim and S. H. Kim, “Pulse wave velocity in atherosclerosis,” Front. Cardiovasc. Med. 6, 41 (2019). [CrossRef]
30. J. Blacher, R. Asmar, S. Djane, G. M. London, and M. E. Safar, “Aortic pulse wave velocity as a marker of cardiovascular risk in hypertensive patients,” Hypertension 33(5), 1111–1117 (1999). [CrossRef]
31. Y. Koji, H. Tomiyama, H. Ichihashi, T. Nagae, N. Tanaka, K. Takazawa, S. Ishimaru, and A. Yamashina, “Comparison of ankle-brachial pressure index and pulse wave velocity as markers of the presence of coronary artery disease in subjects with a high risk of atherosclerotic cardiovascular disease,” Am. J. Cardiol. 94(7), 868–872 (2004). [CrossRef]
32. M. Yambe, H. Tomiyama, Y. Hirayama, Z. Gulniza, Y. Takata, Y. Koji, K. Motobe, and A. Yamashina, “Arterial stiffening as a possible risk factor for both atherosclerosis and diastolic heart failure,” Hypertens. Res. 27(9), 625–631 (2004). [CrossRef]
33. A. Y. Alarhabi, M. S. Mohamed, S. Ibrahim, T. M. Hun, K. I. Musa, and Z. Yusof, “Pulse wave velocity as a marker of severity of coronary artery disease,” J. Clin. Hypertens. (Hoboken, NJ, U. S.) 11(1), 17–21 (2009). [CrossRef]
34. X. N. Liu, H. Q. Gao, B. Y. Li, M. Cheng, Y. B. Ma, Z. M. Zhang, X. M. Gao, Y. P. Liu, and M. Wang, “Pulse wave velocity as a marker of arteriosclerosis and its comorbidities in Chinese patients,” Hypertens. Res. 30(3), 237–242 (2007). [CrossRef]
35. X. Ding, B. P. Yan, Y. T. Zhang, J. Liu, N. Zhao, and H. K. Tsang, “Pulse transit time based continuous cuffless blood pressure estimation: A new extension and a comprehensive evaluation,” Sci. Rep. 7(1), 11554 (2017). [CrossRef]
36. R. Mukkamala, J. O. Hahn, O. T. Inan, L. K. Mestha, C. S. Kim, H. Töreyin, and S. Kyal, “Toward ubiquitous blood pressure monitoring via pulse transit time: theory and practice,” IEEE Trans. Biomed. Eng. 62(8), 1879–1901 (2015). [CrossRef]
37. B. Gribbin, A. Steptoe, and P. Sleight, “Pulse wave velocity as a measure of blood pressure change,” Psychophysiology 13(1), 86–90 (1976). [CrossRef]
38. J. E. Schwartz, P. U. Feig, and J. L. Izzo Jr, “Pulse Wave Velocities Derived From Cuff Ambulatory Pulse Wave Analysis: Effects of Age and Systolic Blood Pressure,” Hypertension 74(1), 111–116 (2019). [CrossRef]
39. T. M. Bashte, Hydraulic Machines and Drives (Moscow), (1970).
40. J. C. Bramwell and A. V. Hill, “The velocity of pulse wave in man,” Proc. R. Soc. Lond. B 93(652), 298–306 (1922). [CrossRef]
41. V. V. Bonarjee, “Arterial stiffness: a prognostic marker in coronary heart disease. available methods and clinical application,” Front. Cardiovasc. Med. 5, 64 (2018). [CrossRef]
42. FREE onlinewebfonts.com, “Human silhouette with white image of the heart Free Icon,” (freepik, 2015). https://www.onlinewebfonts.com/icon/43199
43. Usman Public School System “Circulatory system” (2014). https://www.slideshare.net/123shirazi/circulatory-system-41659320
44. ostamed.ru “Carpal anatomy” (ostamed, 2019). https://ostamed.ru/travmy/perelom/luchezapyastnaya-kost.html
45. D. Massey, “Cardiac assessment part 2: Auscultation,” Br. J. Card. Nurs. 1(3), 126–130 (2006). [CrossRef]
46. J. Nürnberger, S. Dammer, A. O. Saez, T. Philipp, and R. Schäfers, “Diastolic blood pressure is an important determinant of augmentation index and pulse wave velocity in young, healthy males,” J. Hum. Hypertens. 17(3), 153–158 (2003). [CrossRef]
47. E. J. Kim, C. G. Park, J. Park, S. Suh, C. U. Choi, J. W. Kim, S. Kim, H. E. Lim, S. W. Rha, and H. S. Seo, “Relationship between blood pressure parameters and pulse wave velocity in normotensive and hypertensive subjects: invasive study,” J. Hum. Hypertens. 21(2), 141–148 (2007). [CrossRef]
48. M. J. Murphy, K. M. Tichauer, L. Sun, X. Chen, and T. Y. Lee, “Mean transit time as an index of cerebral perfusion pressure in experimental systemic hypotension,” Physiol. Meas. 32(4), 395–405 (2011). [CrossRef]
49. R. Wang, W. Jia, Z. H. Mao, R. J. Sclabassi, and M. Sun, “Cuff-free blood pressure estimation using pulse transit time and heart rate,” in 2014 12th international conference on signal processing (ICSP), pp. 115–118 (2014).
