Open Access
Add to BibSonomy
Abstract
Twin-core fiber (TCF)-based sensor was proposed for non-invasive vital sign monitoring, including respiration and heartbeat. The TCF was homemade and the corresponding sensor was fabricated by sandwiching single-mode fiber (SMF) on both ends. The offset distance between SMF and TCF was optimized while the length of TCF was identified from preliminary vital sign measurement results. Then, the TCF-based sensor was attached under a mattress to realize non-invasive vital sign monitoring. Both respiration and heartbeat signal can be obtained simultaneously, which is consistent with the reference signals. For further application, post-exercise physiological activitity characterization were realized based on this vital sign monitoring system. In discussion, mode coupling in TCF was analyzed and utilized for curvature sensing with achieved sensitivity as high as 18 nm/m−1, which supported its excellent performance for vital signs monitoring. In conclusion, the TCF-based vital signs monitors can be a promising candidate for healthcare and biomedical applications.
© 2019 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Worldwide aging population, one of the current social crises, places a heavy burden on the individuals and society in terms of caregiving and medical expenses. With corresponding change of modern medicine trend from treatment to prevention, daily assessment and monitoring of the health for the elderly become necessary to prevent and control disease development, especially some chronic and senile diseases [1]. Vital signs, such as respiration, heartbeat, body temperature and blood pressure, are the main health indications of human body functions [2]. Good monitoring of basic vital signs parameters can help to assess the physical health condition or even identify specific diseases in the early stage [3]. Thus, vital signs monitoring for aging people becomes promising and urgent for their health trend analysis and disease prevention. It is worth noted that, according to the world market report of telehealth, chronic obstructive pulmonary disease (COPD) and congestive heart failure (CHF) occupy 49% and 12% among senile diseases, indicating the importance of respiration and heartbeat monitoring [4]. Breathing rate (BR) is a critical vital sign reflecting the physiological function of lungs: inflowing of oxygen and removal of carbon dioxide, and abnormal BR is directly related to many symptoms like asthma and anemia [5]. Heart rate (HR), which represents the cardiac cycles from pumping newly oxygenated blood to pulling back deoxygenated blood through the whole body [6], is widely used to assess human physical and mental states [7]. Measurement of these two parameters are the main focus of the current vital signs monitoring.
One of the most popular vital signs monitoring employs the wearable devices, and various sensing schemes are utilized for such monitoring purpose [8]. Currently, for HR monitoring, there are three main sensing techniques, i.e. based on electrical, optical and pressure signals. Electrocardiography (ECG) is used to access human cardiovascular system via depolarization signal pick-up from heart muscles and HR can be obtained from R wave-to-R wave (R-R) interval of ECG signal. Conventional ECG data acquisition requires 12 leads attached on the skin [9] while currently developed techniques only need two electrodes in a band aid form factor [2]. The method using optical [10] and pressure [11] sensors for HR measurement is called plethysmography, which is based on the distention of arteries and arterioles in the subcutaneous tissue due to heart pumping blood. Light-emitting diode and photodetector used in optical sensors capture the light absorption peaks [12] to obtain HR by calculating interval between two systolic peaks while pressure sensors picked up the pulse signals in the same way to calculate HR [13]. For BR measurement, sensors can respond in two ways: expansion and contraction of chest during breathing and flow of breath. Many conductive and dielectric materials were sandwiched between substrates wrapped tightly around the body, which can detect the chest movement to obtain BR [14]. For breath flow detection, the temperature sensors placed near the nose together with the acoustic sensors on the neck can measure BR with high accuracy [15,16]. In summary, the primary advantages of these wearable devices are capability for continuous vital signs monitoring during normal daily life and also the data can be acquired in any circumstances, which provides an opportunity for pneumonic and cardiovascular performance assessment under various settings. However, all these sensing schemes regardless of HR or BR monitoring, require close contact with the body, which are conspicuous and also inevitably discomfort the users, especially the elderly. Moreover, to monitor HR and BR simultaneously, more than one sensor has to be employed and placed in different locations of body, which is not convenient and user-friendly.
To overcome aforementioned drawbacks of wearable devices, non-invasive vital signs monitoring techniques are desirable and have attracted a lot of attentions from researchers in various fields. Different sensing schemes were proposed in recent years. For example, in 2018, Liu et al proposed to track vital signs by using existing off-the-shelf WiFi signals and developed algorithm making use of the channel information in both time and frequency domain to estimate both HR and BR during sleeping [17]. Another recently-reported wireless sensing scheme is Doppler radar. Nosrati et al realized vital signs monitoring in a short range by using a concurrent dual-beam phased-array Doppler radar and MIMO beamforming technique [18]. In addition, near-field coherent sensing method was proposed for the first time for vital signs monitoring. It only needs passive tags placed at the chest and wrist area, retrieving not only HR and BR but also blood pressure and breath effort through the collected multiplexed far-field backscattering waveforms [19]. These non-invasive vital signs monitoring schemes are comfortable to the users and superior to most wearable devices. However, these systems or techniques are complex and high-cost, which is not ideal for practical applications, especially the daily monitoring. On the other hand, their signal detection is somewhat limited to space and distance, resulting in unstable performance in long-term services.
Optical fiber sensors, owing to its advantages of sensitive, low-cost, light-weight, flexible and stable, have been widely used in a wide range of applications. Optical fiber sensors for breath monitoring is to detect the pressure change induced by displacement of chest when breathing on bed while HR is based on the ballistocardiogram (BCG) characterization, which is the recoil forces of the body in reaction to cardiac ejection of blood into the vasculature [20]. This pressure changes and body micromovement are directly related to BR and HR and can be detected simultaneously as well as non-invasively by optical fiber sensors embedded in a mattress. One kind of optical fiber sensor was proposed recently using the amplified bending loss in optical fibers induced by breath and heartbeat [21]. This sensing scheme can realize HR and BR monitoring, but the fabrication possess is very complicated. Our group proposed previously to use different optical fiber interferometers for vital signs monitoring, such as traditional Mach-Zehnder interferometer (MZI) and novel in-line few-mode fiber (FMF) and multi-core fiber (MCF)-based interferometers [22,23]. Traditional MZI with two arms required fussy fabrication process. Also, it is highly sensitive to ambiance factors, namely temperature, which often introduces additional noise during measurement. This makes it difficult for long-term stable vital signs monitoring. For FMF and MCF, they are not sensitive enough to characterize BCG waveform and to obtain accurate HR.
In this paper, to realize non-invasive, convenient, simultaneous and accurate vital signs monitoring, we propose a novel sensing scheme by using a twin-core optical fiber (TCF) and splicing single mode fibers (SMFs) on both ends of TCF. The offset distance between SMF and TCF as well as the length of TCF are investigated and the optimized options are obtained based on preliminary vital signs monitoring results. In experiment, TCF was packaged under a mattress to achieve non-invasive vital signs monitoring. As a result, thanks to its high sensitivity, both breath and heartbeat signal can be simultaneously obtained. In addition, reference breath and heartbeat signals were collected for comparison, which verified the ability of TCF-based sensor for accurate HR and BR monitoring. For further applications, this TCF-based vital signs monitoring system was utilized to characterize human physical recover process after exercise in terms of both amplitude and frequency on heartbeat and breath. Owing to its advantages of easily fabricated, non-invasive, high-sensitive and accurate signal detection, the TCF-based sensor can be a potential and promising candidate for vital signs monitoring.
2. Fabrication of the twin-core fiber and sensor
Figure 1(a) illustrates SEM photo of the cross-section of the TCF, with cladding diameter of ∼125 µm. One core is located in the center with diameter of ∼6 µm and the other identical off-axis core is positioned ∼10 µm away from the center. The TCF-based sensor structure is shown in Fig. 1(c). Due to the adjacency of two cores, their power exchange will happen, which can be characterized by output spectrum in Fig. 1(b). This kind of power exchange, or coupling in TCF, can be described by supermode theory, which is presented in principle and discussion part. It is worth noted that, due to the imperfection of TCF, the two cores are not identical to each other. This may result in unsatisfied spectrum for sensing. Thus, a slight offset between the SMF and TCF is introduced to achieve better spectrum for later vital signs monitoring experiments.
Fig. 1. Cross-section of twin-core fiber (a) and output spectrum (b) in SMF-TCF-SMF sensor structure(c).
In the input end, the core of a single-mode fiber (SMF) is aligned to the center core of TCF. Then, another SMF was off-axis aligned with TCF in the output end and the optimized offset distance can be obtained by tracking the optical spectrum. The offset distance was tuned continuously by splicer (Fujikura, LZM-100) under end-view operation mode while the spectrum was monitored by using Broadband Light Source (BLS) and Optical Spectrum Analyzer (OSA, Yokogawa 6370). The spectra under different offset distances d were collected and plotted in Fig. 2(a). It can be seen that the extinction ratio (ER) increases with the offset distance until the fringe disappeared due to the over offset between the core in SMF and off-axis core in TCF. To make sure that the sensor works under high sensitivity, we finalized the specific offset distance with an ER of ∼5 dB. The direction of the offset between SMF and TCF and the spectra under different offset distances from around 0 µm to 10 µm are shown in Fig. 2(a). Another parameter under optimization is the length L of TCF. Firstly, we recorded the output spectrum of the sensors with different lengths of TCF and compare them as shown in Fig. 2(b). The free spectrum range (FSR) decreases when the length becomes shorter.
Fig. 2. Spectrum collection under different offset distances between SMF and TCF (a) and length of TCF (b).
To optimize the length of TCF L, the spectral response of vital signs signals needs to be monitored in real-time with acceptable accuracy. Thus, an interrogator (Micron Optics, sm130) is employed and a gold filmed is sputtered on the end of TCF to form the coupling in TCF in a reflective way, which are shown in Fig. 3. The interrogator is composed of a laser source which can realize fast wavelength scanning, a photodetector used to record the response under different wavelength and a circulator. The interrogator has a scanning speed up to 2 kHz and a high resolution of 1 pm, which can realize real-time and accurate wavelength shift monitoring. Based on this, the wavelength shift with the vital signs signals are obtained and shown in Fig. 3. It can be seen that the wavelength ranges from 1551.85 nm to 1552.15 nm within 1 min. Thus, the maximum amount of wavelength shift induced by vital signs signals is ∼0.3 nm, which is supposed to be the quasi-linear region in the spectrum. We choose about one third of FSR as the quasi-linear region, therefore, the desirable FSR is ∼0.9 nm and the length of TCF is ∼2 m.
3. Vital signs monitoring system
The vital signs monitoring system is shown in Fig. 4(a). It is worth noted that the two-core orientation in TCF was labeled firstly using the microscope. Then, the whole TCF was placed under the mattress along the labels in attempt to guarantee the correct as well as uniform bending direction. Then, TCF-based sensor was packaged between plastic substrates as a mat under the mattress. A Tunable Laser Source (TLS) was used as input light while a low-speed photodetector converted the light intensity change induced by breath and heartbeat to electrical signal, which was collected by a Data Acquisition (DAQ) card for data processing. Prior to data acquisition, the wavelength of TLS was scanned firstly and then fixed at point A around 1551 nm, as shown in Fig. 4(b), ensuring that the system works on both sensitive and monotonous condition in following data collection process. As a reference, we used a camera to record the rise-and-fall of chest during breath and a commercial ECG device (Sparkfun, AD8232) to obtain actual BR and HR. When the subject is lying on the bed, the camera and the ECG device as well as the fiber sensing system start to acquire the data at the same time. It is worth noted that the subjects are told to stay still during the experiments to avoid undesired spectrum shift due to body movement.
Fig. 4. Vital signs monitoring experiments, including the setup (a) and initial wavelength scanning process (b).
4. Vital signs monitoring results
Due to the intrinsic difference between HR and BR in terms of frequency, the filtering technique was utilized to retrieve the breath and heartbeat waveform. Breath signal can be easily recovered by using a low-pass filter with a bandwidth of 1 Hz. For heartbeat signal, especially for BCG signal recovery, to preserve the accuracy of time interval detection, it has been investigated that the cut-off frequency of low-pass filter should exceed 25 Hz [24]. Thus, a low-pass filter with bandwidth of 30 Hz was applied on the raw data after the breath signal was filtered out to obtain the BCG waveform. Regarding the reference signal, inhale and exhale moments were collected from video for respiration analysis while ECG signal was recorded for heartbeat.
4.1 HR and BR monitoring
The HR and BR monitoring results were shown in Fig. 5, in which blue and red line represent processed and reference signal respectively for both breath and heartbeat results. It can be seen that, for breath monitoring results in Fig. 5(a), the peaks of our processed signal matched well with the reference. Similarly, for BCG waveform in Fig. 5(b), we can easily locate the J peak, the highest peak in standard BCG waveform for HR calculation [20]. Generally, the interval between consecutive J peaks (J-J interval) was used to evaluate HR and so as the R-R interval in ECG waveform. We collect J-J intervals and R-R intervals and get the results of HR and BR as shown in inset of Fig. 5. Therefore, it was demonstrated that the TCF-based vital signs monitoring system can detect both HR and BR simultaneously.
4.2 Post-exercise physiological activities characterization
Human autonomic nervous system (ANS) is comprised of two subsystems: sympathetic nervous system (SNS) and parasympathetic nervous system (PNS). Physiological activities, such as respiration, heartbeat and perspiration, are controlled by the dynamic interaction between SNS and PNS [25]. The full analysis on these physiological activities can be used to assess the autonomic nervous function, which is of great significance for early prevention and detection of various disorders. Among current non-invasive techniques for different physiological activities characterization, heart rate variability (HRV) analysis plays the key role and has been widely investigated [26]. Static exercise, a non-invasive and convenient protocol for altering the hemodynamics and timing interval of heart, was widely used in HRV analysis experiments with data collected on post-exercise heartbeat signals [27]. In addition, R-R interval time series in BCG signal was officially recommended as standards of HRV analysis and HRV analysis using BCG during post-exercise has been demonstrated successfully [28].
In our work, HRV analysis in time domain during post-exercise using BCG signal obtained from TCF-based vital signs monitoring system was conducted. Following the requirement and process of standard HRV analysis experiments, the subject under test was confirmed to be in good health condition with no records of cardiopulmonary disease and was also asked to abstain from caffeine and alcohol for 24 h before the test. During the test, the subject performed burpee exercise for around 20 times within a minute to increase the BR and HR, and then lay on the bed right after the exercise. Data were collected during 4-mins post-exercise period and then processed following the aforementioned methods. The HRV analysis results are shown in Fig. 6. It can be seen that both amplitude and ratio of heartbeat decreased over time during post-exercise and then recovered to normal states. In addition, we also analyzed the breath signals in the same way and presented the results in Fig. 7. In summary, post-exercise physiological activities analysis, including HRV and breath, were successfully performed using our TCF-based vital signs monitoring system, demonstrating its potential applications for the health condition assessment and disease prevention.
5. Principle and discussion
Based on the coupled mode theory, in a single optical fiber with more than one core, or multi-core fiber (MCF), power exchange happens between these cores. To represent this core-to-core crosstalk, supermode theory was proposed and the theoretical analysis on modes in standard multimode fiber can be applied in the same manner on these supermodes in MCF [29]. Thus, the coupling between supermodes also happens in MCF due to the difference of propagation constants. Such coupling scheme has already been utilized to develop optical fiber sensors with excellent performance [30,31]. As for our TCF, we firstly calculate its supported supermodes using COMSOL, and then discuss its application on curvature sensing, which helps to interpret its excellent performance on vital signs monitoring. The curvature sensing principle, experimental setup and results are shown in Appendix. The developed theoretical model of TCF-based sensor for HR and BR monitoring was discussed firstly as follows.
In twin-core fiber (TCF) with cores identical to each other, the output power of center core after length L can be expressed as follows [32]:
For time-dependent response analysis of breath (B) and heartbeat (H), it is assumed that curvature of TCF embedded under mattress is proportionally dependent on the force induced by motion of chest and heart, which can be expressed as follows:
where ρ is the coefficient, t is the time and F is the force. To conclude, combined with Eq. (5), the real-time optical intensity change due to breath and heartbeat can be expressed as:6. Conclusion
In this paper, we proposed an easy and versatile fiber sensing system based on TCF to monitor respiration and heartbeat simultaneously and accurately. The sensor is based on conventional interferometer structure with TCF sandwiched by SMF. To optimize the sensor parameters, including offset distance between SMF and TCF and length of TCF, we compare the collected spectra and performed preliminary vital signs measurement using TCF Michelson interferometers. The optimized TCF-based sensor was attached under the mattress to realize non-invasive monitoring. As a result, both respiration and heartbeat can be obtained simultaneously using filtering techniques and they also matched well with video-captured respiration and ECG heartbeat reference signals. In addition, post-exercise physiological activities were characterized based on this system, which demonstrated its feasibility for HRV analysis. For discussion, symmetric and antisymmetric supermodes in TCF were simulated using COMSOL and their coupling was utilized for curvature sensing. The sensitivity can achieve as high as 18 nm/m−1, which experimentally interpreted its excellent vital signs monitoring performance. In addition, theoretical analysis was preformed to demonstrate the feasibility of TCF-based sensor on HR and BR monitoring. In conclusion, due to the advantages of easy fabrication, low-cost, simultaneous, non-invasive and accurate vital signs monitoring, the TCF-based sensor is desirable as well as promising for wide range of healthcare applications.
Appendix
Based on the basic parameters in TCF, the mode calculation is performed, and the obtained effective refractive indexes of the symmetric mode and antisymmetric mode are 1.4447 and 1.4458, respectively. Under the TCF length of 1 m, the coupling induced output intensity oscillation as a function of the wavelength is simulated and shown in Fig. 8(a). It can be seen that the spectral period is about 2.5 nm, which corresponds to the experimental spectrum in Fig. 1. We also theoretically discuss the influence of offset distance between SMF and TCF and the length of TCF on the spectrum. The comparison between the simulation results and the experiment results from Fig. 2 are shown in Fig. 9. It can be seen that the simulation results agree well with the experiment with acceptable errors.
Fig. 8. The symmetric and antisymmetric supermode supported in TCF and the simulated spectrum (a) and schematic of bending TCF (b).
TCF has already been demonstrated to be highly sensitive the curvature in both theoretical [33] and experimental [34] way. We follow the bending TCF model analysis and also conduct the curvature sensing experiment on our TCF-based sensor. The schematic of bending TCF is shown in Fig. 8(b), in which R is the bending radius and x represents the bending direction. When TCF is under bending, one core is in tension while the other one is in compression. The refractive index after bending can be expressed by
where C = 1/R is the curvature, n(x) and n’(x) are refractive index profile when TCF is straight and bent, respectively [33]. Thus, bending will induce the change on the refractive index of these two cores in opposite direction. Since supermode is demonstrated to be highly sensitive to unidentical refractive index change of cores [29], under different curvatures, ns and na will change dramatically, leading to large wavelength shift in optical spectrum.In experiment, the experimental setup used to characterize the bending is shown in Fig. 10(a), in which the BLS and OSA are used for the spectrum monitoring. Left translation stage moves in x direction while the right one is fixed. Prior to spectrum collection, we labeled the two-core orientation from microscopy firstly and rotated two holders to make the bending direction and two cores in one plane, as shown in Fig. 10(a). The bending radius R can be calculated from the movement distance and the length of TCF L0. The curvature sensing results are illustrated in Fig. 10(b). It can be seen that, sensitivity of this TCF-based sensor can achieve as high as 18 nm/m−1. Such high bending dependence proves the feasibility of the vital signs monitoring based on the similar sensing scheme introduced above.
Funding
Hong Kong Polytechnic University (1-ZVGB, 1-ZVHA); Research Grants Council, University Grants Committee (15211317).
Disclosures
The authors declare that there are no conflicts of interest related to this article.
References
1. S. G. Alonso, I. de la Torre Díez, and B. G. Zapiraín, “Predictive, Personalized, Preventive and Participatory (4P) Medicine Applied to Telemedicine and eHealth in the Literature,” J. Med. Syst. 43(5), 140 (2019). [CrossRef]
2. Y. Khan, A. E. Ostfeld, C. M. Lochner, A. Pierre, and A. C. Arias, “Monitoring of Vital Signs with Flexible and Wearable Medical Devices,” Adv. Mater. 28(22), 4373–4395 (2016). [CrossRef]
3. M. Weenk, M. Koeneman, T. H. van de Belt, L. J. L. P. G. Engelen, H. van Goor, and S. J. H. Bredie, “Wireless and continuous monitoring of vital signs in patients at the general ward,” Resuscitation 136, 47–53 (2019). [CrossRef]
4. F. Zhang, Y. Yu, and J. Zhong, “Research Status and Development Prospects of Human Vital Signs Monitoring Clothing,” IOP Conf. Ser.: Earth Environ. Sci. 233(4), 042031 (2019). [CrossRef]
5. B. Bochner and W. Busse, “Respiratory System, Part Two: Allergy and Asthma,” J. Allergy Clin. Immunol. 115(5), 953–959 (2005). [CrossRef]
6. A. Stephenson, J. W. Adams, and M. Vaccarezza, “The vertebrate heart: an evolutionary perspective,” J. Anat. 231(6), 787–797 (2017). [CrossRef]
7. G. A. Fonzo and A. Etkin, “Brain Connectivity Reflects Mental and Physical States in Generalized Anxiety Disorder,” Biol. Psychiatry 80(10), 733–735 (2016). [CrossRef]
8. Z. Wang, Z. Yang, and T. Dong, “A Review of Wearable Technologies for Elderly Care that Can Accurately Track Indoor Position, Recognize Physical Activities and Monitor Vital Signs in Real Time,” Sensors 17(2), 341 (2017). [CrossRef]
9. Z. C. Haberman, R. T. Jahn, R. Bose, H. Tun, J. S. Shinbane, R. N. Doshi, P. M. Chang, and L. A. Saxon, “Wireless Smartphone ECG Enables Large-Scale Screening in Diverse Populations,” J. Cardiovasc. Electrophysiol. 26(5), 520–526 (2015). [CrossRef]
10. C. M. Lochner, Y. Khan, A. Pierre, and A. C. Arias, “All-organic optoelectronic sensor for pulse oximetry,” Nat. Commun. 5(1), 5745 (2014). [CrossRef]
11. J. Harju, A. Tarniceriu, J. Parak, A. Vehkaoja, A. Yli-Hankala, and I. Korhonen, “Monitoring of heart rate and inter-beat intervals with wrist plethysmography in patients with atrial fibrillation,” Physiol. Meas. 39(6), 065007 (2018). [CrossRef]
12. B. Desai and U. Chaskar, “Comparison of optical sensors for non-invasive hemoglobin measurement,” in 2016 International Conference on Electrical, Electronics, and Optimization Techniques (ICEEOT), 2211–2214 (2016).
13. G. Schwartz, B. C. K. Tee, J. Mei, A. L. Appleton, D. H. Kim, H. Wang, and Z. Bao, “Flexible polymer transistors with high pressure sensitivity for application in electronic skin and health monitoring,” Nat. Commun. 4(1), 1859 (2013). [CrossRef]
14. M. Folke, L. Cernerud, M. Ekström, and B. Hök, “Critical review of non-invasive respiratory monitoring in medical care,” Med. Biol. Eng. Comput. 41(4), 377–383 (2003). [CrossRef]
15. P. Corbishley and E. Rodriguez-Villegas, “Breathing Detection: Towards a Miniaturized, Wearable, Battery-Operated Monitoring System,” IEEE Trans. Biomed. Eng. 55(1), 196–204 (2008). [CrossRef]
16. O. Mimoz, T. Benard, A. Gaucher, D. Frasca, and B. Debaene, “Accuracy of respiratory rate monitoring using a non-invasive acoustic method after general anaesthesia,” Br. J. Anaesth. 108(5), 872–875 (2012). [CrossRef]
17. J. Liu, Y. Chen, Y. Wang, X. Chen, J. Cheng, and J. Yang, “Monitoring Vital Signs and Postures During Sleep Using WiFi Signals,” IEEE Internet Things J. 5(3), 2071–2084 (2018). [CrossRef]
18. M. Nosrati, S. Shahsavari, S. Lee, H. Wang, and N. Tavassolian, “A Concurrent Dual-Beam Phased-Array Doppler Radar Using MIMO Beamforming Techniques for Short-Range Vital-Signs Monitoring,” IEEE Trans. Antennas Propag. 67(4), 2390–2404 (2019). [CrossRef]
19. X. Hui and E. C. Kan, “Monitoring vital signs over multiplexed radio by near-field coherent sensing,” Nat. Electron. 1(1), 74–78 (2018). [CrossRef]
20. O. T. Inan, P. F. Migeotte, K.-S. Park, M. Etemadi, K. Tavakolian, R. Casanella, J. Zanetti, J. Tank, I. Funtova, G. K. Prisk, and M. D. Rienzo, “Ballistocardiography and Seismocardiography: A Review of Recent Advances,” IEEE J. Biomed. Health Inform. 19(4), 1414–1427 (2015). [CrossRef]
21. X. Yang, Z. Chen, C. S. M. Elvin, L. H. Y. Janice, S. Ng, J. Teo, and R. Wu, “Textile Fiber Optic Microbend Sensor Used for Heartbeat and Respiration Monitoring,” IEEE Sens. J. 15(2), 757–761 (2015). [CrossRef]
22. S. Chen, Z. Huang, F. Tan, T. Yang, J. Tu, and C. Yu, “Vital signs monitoring using few-mode fiber-based sensors,” in Optoelectronic Devices and Integration VII in Photonics Asia 2018, 10814 (2018).
23. S. Chen, F. Tan, Z. Huang, T. Yang, J. Tu, and C. Yu, “Non-invasive smart monitoring system based on multi-core fiber optic interferometers,” in 2018 Asia Communications and Photonics Conference (ACP), 1–3 (2018).
24. J. H. Shin and K. S. Park, “HRV analysis and blood pressure monitoring on weighing scale using BCG,” in 2012 Annual International Conference of the IEEE Engineering in Medicine and Biology Society, 3789–3792 (2012).
25. J. H. Shin, S. H. Hwang, M. H. Chang, and K. S. Park, “Heart rate variability analysis using a ballistocardiogram during Valsalva manoeuvre and post exercise,” Physiol. Meas. 32(8), 1239–1264 (2011). [CrossRef]
26. Y. Zhong, K. Jan, K. H. Ju, and K. H. Chon, “Quantifying cardiac sympathetic and parasympathetic nervous activities using principal dynamic modes analysis of heart rate variability,” Am. J. Physiol. Circ. Physiol. 291(3), H1475–H1483 (2006). [CrossRef]
27. R. González-Camarena, S. Carrasco-Sosa, R. Román-Ramos, M. J. Gaitán-González, V. Medina-Bañuelos, and J. Azpiroz-Leehan, “Effect of static and dynamic exercise on heart rate and blood pressure variabilities,” Med. Sci. Sports Exercise 32(10), 1719–1728 (2000). [CrossRef]
28. A. J. Camm, M. Malik, J. T. Bigger, G. Breithardt, S. Cerutti, R. J. Cohen, P. Coumel, E. L. Fallen, H. L. Kennedy, R. E. Kleiger, L. Federico, M. Alberto, A. J. Moss, J. N. Rottman, G. Schmidt, P. J. Schwartz, and D. H. Singer, “Heart rate variability. Standards of measurement, physiological interpretation, and clinical use. Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology,” Eur. Heart J. 17(3), 354–381 (1996). [CrossRef]
29. L. Szostkiewicz, M. Napierala, A. Ziolowicz, A. Pytel, T. Tenderenda, and T. Nasilowski, “Cross talk analysis in multicore optical fibers by supermode theory,” Opt. Lett. 41(16), 3759 (2016). [CrossRef]
30. J. E. Antonio-Lopez, Z. S. Eznaveh, P. LiKamWa, A. Schülzgen, and R. Amezcua-Correa, “Multicore fiber sensor for high-temperature applications up to 1000(C,” Opt. Lett. 39(15), 4309–4312 (2014). [CrossRef]
31. J. Villatoro, O. Arrizabalaga, G. Durana, I. Ocariz, E. Lopze, J. Zubia, A. Schulzgen, and R. Amezcua-Correa, “Accurate strain sensing based on super-mode interference in strongly coupled multi-core optical fibres,” Sci. Rep. 7(1), 4451 (2017). [CrossRef]
32. P. Rugeland and W. Margulis, “Revisiting twin-core fiber sensors for high-temperature measurements,” Appl. Opt. 51(25), 6227–6232 (2012). [CrossRef]
33. G. L. Yin, S. Q. Lou, W. L. Lu, and X. Wang, “A high-sensitive fiber curvature sensor using twin core fiber-based filter,” Appl. Phys. B: Lasers Opt. 115(1), 99–104 (2014). [CrossRef]
34. Y. Wu, L. Pei, W. Jin, Y. Jiang, Y. Yang, Y. Shen, and S. Jian, “Highly sensitive curvature sensor based on asymmetrical twin core fiber and multimode fiber,” Opt. Laser Technol. 92, 74–79 (2017). [CrossRef]
