Non-invasive vein waveform analysis for evaluating subject

Abstract

To solve the problem with HRV normally derived from an electrocardiogram (ECG) that a contact electrode generally needs to be used, which sometimes poses inconvenience.SOLUTION: An exemplary method includes: (a) generating a signal indicating vibration generated from a blood vessel of a subject through a sensor of a computing device, the vibration indicating a heart beat and / or respiration of the subject; (b) determining a first time elapsing between each pair of continuous heart beats shown by the vibration and / or a second time elapsing between each pair of continuous respirations shown by the vibration by using the signal; and (c) determining heart rate variations of the subject using the determined first elapsing time and / or determining respiration variations of the subject using the determined second elapsing time.SELECTED DRAWING: Figure 4

Description

Technical Field

[0001] Cross - Reference to Related Applications This application claims priority to U.S. Provisional Patent Application No. 62 / 727,735, filed on September 6, 2018, the content of which is hereby incorporated by reference in its entirety.

Background Art

[0002] Unless otherwise specifically indicated herein, the content described in this section is not prior art to the claims in this application and is not admitted to be prior art by virtue of being included in this section.

[0003] Heart rate variability (HRV) can be used to evaluate the autonomic nervous function of a subject. More specifically, HRV can be used to evaluate the responses of the sympathetic (stress) and parasympathetic (anti - stress) nerves of the nervous system. One problem that may be encountered is that the frequency ranges of HRV representing the sympathetic and parasympathetic responses can overlap to some extent. HRV is usually derived from an electrocardiogram (ECG), but generally requires the use of contact electrodes, which can be inconvenient. Therefore, there is a need for a more convenient means for evaluating HRV and for better distinguishing the frequency ranges of HRV corresponding to the sympathetic and parasympathetic responses of the nervous system, respectively.

Summary of the Invention

[0004] In one example, the method includes: (a) generating, via a sensor of a computing device, a signal representative of vibrations generated from a subject's blood vessels, the vibrations indicating the subject's heartbeat and / or respiration; (b) using the signal to determine a first time elapsed between respective pairs of successive heartbeats indicated by the vibrations and / or a second time elapsed between respective pairs of successive respirations indicated by the vibrations; and (c) using the determined first elapsed time to determine the subject's heart rate variability and / or using the determined second elapsed time to determine the subject's respiratory rate variability.

[0005] In another example, a computing device includes one or more processors, a sensor, and a computer-readable medium storing instructions that, when executed by the one or more processors, cause the computing device to perform functions. The functions include: (a) generating, via the sensor, a signal representative of vibrations generated from a subject's blood vessels, the vibrations indicating the subject's heartbeat and / or respiration; (b) using the signal to determine a first time elapsed between respective pairs of successive heartbeats indicated by the vibrations and / or a second time elapsed between respective pairs of successive respirations indicated by the vibrations; and (c) using the determined first elapsed time to determine the subject's heart rate variability and / or using the determined second elapsed time to determine the subject's respiratory rate variability.

[0006] In yet another example, a non-transitory computer-readable medium stores instructions that, when executed by a computing device including a sensor, cause the computing device to perform functions. The functions include: (a) generating, via the sensor, a signal representative of vibrations generated from a blood vessel of a subject, the vibrations indicating the subject's heartbeat and / or respiration; (b) using the signal to determine a first time elapsed between respective pairs of consecutive heartbeats indicated by the vibrations and / or a second time elapsed between respective pairs of consecutive respirations indicated by the vibrations; and (c) using the determined first elapsed time to determine a variation in the subject's heart rate and / or using the determined second elapsed time to determine a variation in the subject's respiration rate. These and other aspects, advantages, and alternatives will become apparent to those of ordinary skill in the art by reading the following detailed description with appropriate reference to the accompanying drawings. Further, the summary and other descriptions and figures provided herein are intended to illustrate the invention by way of example only and, thus, it is to be understood that many variations are possible. BRIEF DESCRIPTION OF THE DRAWINGS

[0007]

Figure 1

Figure 2

Figure 3A

Figure 3B

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14

Figure 15

Figure 16

DETAILED DESCRIPTION OF THE INVENTION

[0008] As described above, there is a need for a more convenient means for evaluating HRV and for better distinguishing the frequency ranges of HRV corresponding to the sympathetic and parasympathetic responses of the nervous system, respectively. The embodiments disclosed herein not only help address these issues, but can also help better obtain fluctuations in the respiratory rate of a subject.

[0009] For example, a computing device can include a (e.g., piezoelectric) sensor wearable on a subject's wrist that can generate a signal representative of vibrations generated from the subject's blood vessels. In this context, the vibrations indicate the subject's heartbeat and / or respiration. The computing device can use the signal to determine a first time elapsed between each pair of consecutive heartbeats indicated by the vibrations and / or a second time elapsed between each pair of consecutive respirations indicated by the vibrations. The computing device can then use the determined first elapsed time to determine the subject's heart rate variability and / or use the determined second elapsed time to determine the subject's respiratory rate variability.

[0010] FIG. 1 is a simplified block diagram of an exemplary computing device 100 that can perform various operations and / or functions, such as those described in the present disclosure. The computing device 100 can be, among other possibilities, a mobile phone, a tablet computer, a laptop computer, a desktop computer, a wearable computing device (e.g., in the form of a wristband).

[0011] Computing device 100 includes one or more processors 102, a computer-readable medium 104, a communication interface 106, a user interface 108, a display 110, and a sensor 112. These components, and other possible components, can be connected to each other (or to another device or system) via a connection mechanism 114, which represents a mechanism that facilitates communication between two or more devices or systems. Thus, the connection mechanism 114 can be a simple mechanism such as a cable or a system bus, or a relatively complex mechanism such as a packet-based communication network (e.g., the Internet). In some examples, the connection mechanism can include an intangible medium (e.g., if the connection is wireless).

[0012] One or more processors 102 can include a general-purpose processor (e.g., a microprocessor) and / or a dedicated processor (e.g., a digital signal processor (DSP)). In some examples, computing device 100 can include two or more processors to perform the functions described herein.

[0013] The computer-readable medium 104 may include one or more volatile, non-volatile, removable, and / or non-removable storage components such as magnetic, optical, or flash storage, and / or may be integrated, in whole or in part, with one or more processors 102. Thus, the computer-readable medium 104 can take the form of a non-transitory computer-readable storage medium that stores thereon program instructions (e.g., compiled or non-compiled program logic and / or machine language), which, when executed by one or more processors 102, cause the program instructions to cause the computing device 100 to perform one or more operations and / or functions such as those described in this disclosure. Such program instructions can define, and / or be part of, an individual software application. In some examples, the computing device 100 can execute the program instructions in response to receiving input from, for example, the communication interface 106 and / or the user interface 108. The computer-readable medium 104 can also store other types of data such as those described in this disclosure.

[0014] The communication interface 106 can enable the computing device 100 to connect and / or communicate with another device or system according to one or more communication protocols. The communication interface 106 can be a wired interface such as an Ethernet interface or a high-definition serial digital interface (HD-SDI). The communication interface 106 can additionally or alternatively include a wireless interface such as a cellular or WI-FI interface. The connection provided by the communication interface 106 can be a direct connection or an indirect connection, the latter being a connection that passes through and / or traverses one or more entities such as a router, switch, or other network device. Similarly, a transmission to or from the communication interface 106 can be a direct transmission or an indirect transmission.

[0015] The user interface 108 can, if applicable, facilitate the interaction between the computing device 100 and the user of the computing device 100. To that end, the user interface 108 can include input components such as a keyboard, keypad, mouse, touch-sensitive and / or presence-sensitive pad or display, microphone, camera, etc., and output components such as a display device, speaker, and / or haptic feedback system (which can be combined with, for example, a touch-sensitive and / or presence-sensitive panel). More generally, the user interface 108 can include any hardware and / or software components that facilitate the interaction between the computing device 100 and the user of the computing device 100.

[0016] In a further aspect, the computing device 100 includes a display 110. The display 110 can be any type of graphic display. As such, the display 110 can vary in size, shape, and / or resolution. Further, the display 110 can be a color display or a monochrome display.

[0017] The sensor 112 can take the form of a piezoelectric sensor, pressure sensor, capacitance sensor, strain sensor, force sensor, light wavelength selective reflectance or absorbance measurement system, tonometer, ultrasonic probe, plethysmograph, or pressure transducer. Other examples are possible. The sensor 112 is configured to detect vibrations arising from the blood vessels of a subject, as further described herein.

[0018] Throughout this book, the phrase "sensors of a computing device" can refer to embodiments where the sensors are integrated with and / or are in a wired connection with the computing device. Further, "sensors of a computing device" can refer to embodiments where the sensors have a permanent, non-permanent, and / or intermittent wireless connection to the computing device. In some embodiments, the term "computing device" can generically refer to a "master" device such as a tablet computer, laptop computer, desktop computer, or mobile phone, and "slave devices" such as sensors having a wireless connection to the master device.

[0019] FIG. 2 shows an embodiment of a computing device 100 and a sensor 112. In FIG. 2, the sensor 112 takes the form of a wearable wristband worn by a human subject, and the computing device 100 takes the form of a mobile phone. The sensor 112 can detect vibrations generated from the blood vessels (e.g., arteries or veins) of the subject's wrist and wirelessly transmit (e.g., via Bluetooth®) a signal representing the detected vibrations to the computing device 100. The computing device 100 can receive the signal for further processing as further described herein.

[0020] FIG. 3A shows another embodiment of the computing device 100. In FIG. 3A, the computing device 100 is communicatively coupled to the sensor 112 via a wired connection.

[0021] FIG. 3B shows an embodiment of the sensor 112 in the form of a wristband.

[0022] FIG. 4 is a block diagram of a method 400 executable by and / or via use of the computing device 100.

[0023] In block 402, method 400 includes generating a signal representative of vibrations originating from a subject's blood vessels via a sensor of a computing device. In this context, the vibrations indicate the subject's heartbeat and / or respiration.

[0024] For example, computing device 100 can detect vibrations originating from a subject's blood vessels (e.g., vein wall or artery wall) via sensor(s) 112. Sensor(s) 112 can be positioned, for example, proximate to a subject's peripheral vein or artery (e.g., wrist) and detect vibrations originating from the peripheral vein or artery. The computing device 100 can then generate a signal representative of the detected vibrations.

[0025] The vibrations can be generated by the fluid flowing through the blood vessels, can be generated by the wall tension of the blood vessels, or can be generated by the contraction or relaxation of the blood vessels in response to (e.g., physiologically) the fluid flowing through the blood, or by the subject's respiration.

[0026] In a particular example, sensor(s) 112 can be secured (e.g., via a VELCRO™ strap) to the subject's skin over or near the blood vessel (see FIG. 2). In this case, the blood vessel can be a forearm vein. Sensor(s) 112 can detect vibrations caused by the blood flow through the blood vessel when the vibrations are conducted through tissue such as the subject's skin.

[0027] The subject can be a human, but other animals are also possible. When sensor(s) 112 detect the vibrations, the subject can breathe spontaneously, for example, without the aid of a ventilator or with the aid of a ventilator.

[0028] Referring to FIG. 5, the signal labeled "NIVA" (non-invasive venous waveform analysis) represents the signal generated by a sensor such as sensor 112 as described above. The NIVA signal can represent the degree of displacement of the sensor caused by the vibration of the blood vessel over time. The signal labeled "ECG" was obtained from the same subject using an electrocardiogram simultaneously with the acquisition of the NIVA signal. The ECG signal represents the voltage representing various stages of the subject's heartbeat detected on the subject's skin over time. The signal labeled "RES" was obtained from the same subject wearing an expandable chest band simultaneously with the acquisition of the NIVA signal. The RES signal represents the expansion and contraction of the subject's chest due to breathing over time. NIVA (e.g., the techniques disclosed herein) has been shown to be capable of obtaining a signal suitable for accurately determining the respiratory rate of a subject, similar to the signals obtained using, for example, an ECG or an expandable chest band.

[0029] In block 404, the method includes using the signal to determine a first time elapsed between each pair of consecutive heartbeats indicated by the vibration and / or a second time elapsed between each pair of consecutive breaths indicated by the vibration. For example, computing device 100 can use the NIVA signal to determine a first time elapsed between each pair of consecutive heartbeats indicated by the vibration and / or a second time elapsed between each pair of consecutive breaths indicated by the vibration, as described below.

[0030] As a first step, computing device 100 can low-pass filter the NIVA signal. For example, referring to FIG. 6, the raw NIVA signal 602, the low-pass filtered NIVA signal 604, and the amplified NIVA signal 606 are shown. The raw NIVA signal 602 represents a signal generated by computing device 100. The low-pass filtered NIVA signal 604 represents the output of computing device 100 that low-pass filters the raw NIVA signal 602. The amplified NIVA signal 606 represents the output of computing device 100 that amplifies the low-pass filtered NIVA signal 604. The amplitudes of the signals 602, 604, and 606 shown are not necessarily shown to scale.

[0031] Referring to FIG. 7, computing device 100 can then apply the amplified NIVA signal 606 to a low-pass differentiator filter to generate signal 608. As will be described below, either signal 606 or signal 608 can be used by computing device 100 for peak identification or maximum gradient identification. Computing device 100 can also identify the full cycle of the NIVA signal in other ways. Generally, each peak (e.g., full cycle) of signal 606 or signal 608 can be considered to represent a single heartbeat of the subject. Additionally, low-frequency information can be decomposed from the signal to obtain the respiration rate and variations in the respiration rate.

[0032] In some embodiments, computing device 100 can identify peaks (e.g., maxima) 710A, 710B, 710C, 710D, 710E, 710F, 710G, 710H, 710I, 710J, 710K, 710L, 710M, and 710N of signal 606. (In other embodiments, computing device 100 can also identify minima of signal 606). Computing device 100 can identify the peaks, for example, using an adaptive threshold. Next, computing device 100 can determine the elapsed time between successive peaks 710A and 710B, between successive peaks 710B and 710C, between successive peaks 710C and 710D, etc., until it determines the elapsed time between successive peaks 710M and 710N. As will be described below with respect to block 406, these times elapsed between successive heartbeats can be further used by computing device 100. Using similar techniques, the respiratory rate and fluctuations in the respiratory rate can be obtained.

[0033] In a similar manner, computing device 100 can identify points of maximum (e.g., positive) slope of signal 606, e.g., points 715A, 715B, 715C, 715D, 715E, 715F, 715G, 715H, 715I, 715J, 715K, 715L, 715M, and 715N. (In other embodiments, computing device 100 can also identify points of maximum negative slope of signal 606). Next, computing device 100 can determine the elapsed time between successive points 715A and 715B, between successive points 715B and 715C, between successive points 715C and 715D, etc., until it determines the elapsed time between successive points 715M and 715N. As will be described below with respect to block 406, these times elapsed between successive heartbeats can be further used by computing device 100. Using similar techniques, the respiratory rate and fluctuations in the respiratory rate can be obtained.

[0034] In some embodiments, computing device 100 can identify peaks (e.g., maxima) 720A, 720B, 720C, 720D, 720E, 720F, 720G, 720H, 720I, 720J, 720K, 720L, 720M, and 720N of signal 608. (In other embodiments, computing device 100 can also identify minima of signal 608). Computing device 100 can identify the peaks, for example, using an adaptive threshold. Next, computing device 100 can determine the elapsed time between consecutive peaks 720A and 720B, between consecutive peaks 720B and 720C, between consecutive peaks 720C and 720D, etc., until it determines the elapsed time between consecutive peaks 720M and 720N. As will be described below with respect to block 406, these times elapsed between consecutive heartbeats can be further used by computing device 100. Similar techniques can be used to obtain the respiratory rate and fluctuations in the respiratory rate.

[0035] In block 406, the method includes determining the subject's heart rate variability using the determined first time elapsed between each pair of consecutive heartbeats and / or determining the subject's respiratory rate variability using the determined second time elapsed. For example, computing device 100 can determine the subject's heart rate variability using the elapsed time between (i) pairs of consecutive peaks among peaks 710A - N, (ii) pairs of consecutive points among points 715A - 715N of maximum gradient, and / or (iii) pairs of consecutive peaks among peaks 720A - N. Typically, computing device 100 uses a NIVA signal such as signal 606 or signal 608 (or an unprocessed NIVA signal or a NIVA signal processed by other means) to calculate, for example, the variance or standard deviation of all elapsed times between consecutive heartbeats, such as between peaks of the NIVA signal, between minima or maxima of the NIVA signal, and / or between consecutive points of the NIVA signal having the absolute maximum of the gradient. Similar techniques can be used to obtain the respiratory rate and the variability of the respiratory rate.

[0036] FIG. 8 is a scatter plot showing the elapsed time between consecutive heartbeats as a function of time. The star-shaped points represent the elapsed time 802 when determined by analyzing an ECG signal (e.g., the ECG signal of FIG. 5) (e.g., by computing device 100). The x-shaped points represent the elapsed time 804 when determined by analyzing a NIVA signal (e.g., NIVA signals 602, 604, 606, or 608) (e.g., by computing device 100). FIG. 8 shows that the elapsed time between consecutive heartbeats for this particular subject was in the range of about 830 milliseconds to 1050 milliseconds over a duration of about 150 seconds. The time between the subject's heartbeats increased slightly during that period. Similar data collection techniques can be used to obtain the variability of the respiratory rate.

[0037] Thus, block 406 can include the computing device 100 calculating all variances of the elapsed time 804, perhaps using the following equation (1). In the context of equation (1), s 2 is equal to the variance of the elapsed time between successive heartbeats of the subject, i refers to each period in the elapsed time dataset, and t M refers to the arithmetic mean of all the elapsed times between successive heartbeats of the subject, and n refers to the amount of elapsed time in the dataset. The calculated variance represents a proxy for the heart rate variability (HRV) of the subject. In other embodiments, the standard deviation of all the elapsed times between successive heartbeats of the subject can be calculated and used as a surrogate for the HRV of the subject. Other quantitative metrics can also be used as proxies for HRV.

[0038]

Number

[0039] In various embodiments, it may be useful to use a dataset representing heartbeats that all occur within a particular duration, for example, in the range of 1 second to 10 seconds, 2 seconds to 8 seconds, or 2.5 seconds to 6 seconds. These time ranges typically correspond to the duration of the subject's respiration. The NIVA technique itself can be used to determine the actual respiration rate of the subject, and a time window can be selected based on the determined respiration rate. By focusing on periods that somewhat closely correspond to the actual respiration rate of the subject, data that better represents the subject's parasympathetic response can be obtained. An enhanced parasympathetic response generally corresponds to higher HRV in the frequency range of about 0.15 to 0.4 Hz.

[0040] Figure 9 shows the power spectral density of the NIVA signal. For example, computing device 100 can obtain the power spectral density of NIVA signals such as signals 602, 604, 606, or 608 by performing a Fourier transform on the NIVA signal. Next, computing device 100 can determine the lowest frequency of the power spectral density at which there is a peak exceeding a predetermined power threshold. For example, in Figure 9, the predetermined threshold power can be set to 1x10 -5 in arbitrary units, where f0 is the lowest frequency of the power spectral density at which there is a peak exceeding the threshold. Thus, the peak "f0" corresponds to the respiratory rate of the subject and, in this case, is approximately 0.2 Hz. The peak "f1" corresponds to the heart rate of the subject and, in this case, is approximately 1.45 Hz.

[0041] In this example, the respiratory rate of 0.2 Hz of the subject corresponds to a reciprocal time window of 5 seconds. Thus, computing device 100 can generate the NIVA signal for at least 5 seconds, or perhaps 10 seconds, and calculate the variance or standard deviation of the elapsed time between consecutive heartbeats within that time window. In another embodiment, computing device 100 can generate NIVA data corresponding to several minutes and calculate several variances, each variance corresponding to a separate time window lasting approximately 5 - 10 seconds (or another duration). The time window corresponding to the respiratory rate of the subject (e.g., several seconds) should have information that best represents the parasympathetic response, but performing measurements several times over several minutes can help improve accuracy. Such data can be used to evaluate the response of the subject's parasympathetic nervous system based on the determined heart rate variability. The variance of data corresponding to the elapsed time between consecutive heartbeats over a larger time window (e.g., minutes) can be used to evaluate the response of the subject's sympathetic nervous system based on the determined heart rate variability.

[0042] The methods disclosed herein can evaluate the effectiveness of treatment protocols such as painkillers, antidepressants, yoga, oxytocin administration, anesthesia administration, sedative administration, anxiolytic administration, behavioral therapy intervention, sympathetic corticoid administration, alpha and beta antagonist administration, environmental intervention (e.g., subject exposure to reduced light compared to a control, or reduced noise compared to a control), music therapy, art therapy, virtual reality therapy, pet therapy, etc.

[0043] Using the disclosed methods, post-traumatic stress disorder, anxiety disorders, pain, depression, delirium, severe illness, postoperative stress, traumatic brain injury, cancer, after myocardial infarction, complications caused by local anesthesia or conscious sedation, burns, or complications of childbirth can be treated or diagnosed.

[0044] For example, using the disclosed methods, the effectiveness of treatments for PTSD such as psychotherapy or antidepressants can be evaluated. Such antidepressants can also be used to help with the symptoms of depression and anxiety disorders. They can also help improve sleep disorders and concentration. Such drugs can include selective serotonin reuptake inhibitor (SSRI) therapy, sertraline (Zoloft) and paroxetine (Paxil), or prazosin.

[0045] Using the disclosed methods, the effectiveness of treatments for anxiety disorders such as psychotherapy and antidepressants, including drug therapies of the selective serotonin reuptake inhibitor (SSRI) and serotonin-norepinephrine reuptake inhibitor (SNRI) classes, can also be evaluated. Examples of antidepressants used in the treatment of generalized anxiety disorder include escitalopram (Lexapro), duloxetine (Cymbalta), venlafaxine (Effexor XR) and paroxetine (Paxil, Pexeva), buspirone, or benzodiazepine.

[0046] Using the disclosed method, the effectiveness of treating depression can also be evaluated, such as SSRIs including citalopram (Celexa), escitalopram (Lexapro), fluoxetine (Prozac), paroxetine (Paxil, Pexeva), sertraline (Zoloft), and vilazodone (Viibryd), or serotonin-norepinephrine reuptake inhibitors (SNRIs). Examples of SNRIs include duloxetine (Cymbalta), venlafaxine (Effexor XR), desvenlafaxine (Pristiq, Khedezla), and levomilnacipran (Fetzima). Using the disclosed method, the effectiveness of bupropion (Wellbutrin XL, Wellbutrin SR, Aplenzin, Forfivo XL), mirtazapine (Remeron), nefazodone, trazodone, and vortioxetine (Trintellix) can also be evaluated. Using the disclosed method, the effectiveness of tricyclic antidepressants such as imipramine (Tofranil), nortriptyline (Pamelor), amitriptyline, doxepin, trimipramine (Surmontil), desipramine (Norpramin), and protriptyline (Vivactil) can also be evaluated. Using the disclosed method, the effectiveness of monoamine oxidase inhibitors (MAOIs) such as tranylcypromine (Parnate), phenelzine (Nardil), and isocarboxazid (Marplan) selegiline (Emsam) can also be evaluated.

[0047] Figure 10 shows the general relationship between the signal powers of the ECG and NIVA obtained simultaneously from a subject. As shown, at low frequencies (PLF) (e.g., 0.04 - 0.15 Hz), the signal-to-noise ratios (SNRs) of the NIVA signal and the ECG signal are similar, but at high frequencies (PHF) (e.g., 0.15 - 0.4 Hz), NIVA has a significantly higher SNR.

[0048] Figure 11 shows data corresponding to subjects collected before and after 15 subjects performed yoga practice. The data shows higher HRV within the frequency range corresponding to the subjects' respiratory rate (e.g., 0.04 - 0.15 Hz) after the subjects performed yoga exercises. More specifically, in this experiment, the frequency range was defined as the detected respiratory rate of each subject + / - 0.025 Hz. This data indicates that yoga practice may lead to the activation of the parasympathetic nervous system of the subjects.

[0049] The following text contains further discussion of the experimental results.

[0050] Figure 12 shows the relationship between the peak power of the respiratory rate and the tidal volume in a pig model. As shown, the peak signal power generally correlates with the tidal volume.

[0051] Figure 13 shows the NIVA waveform (upper panel) with pulse detection (circles) including the incident (cross) and the corresponding pulse interval series (lower panel) with outliers' intervals removed.

[0052] Figure 14 shows the NN interval series (left panel) before and after a yoga session and the corresponding power spectral density (right panel).

[0053] Figure 15 shows the transient PRV parameters in the recordings before and after yoga (* indicates a significant difference (p < 0.05)).

[0054] Figure 16 shows the spectral and PRV parameters of LLP (SD12, lags 5 and 7) in the recordings before and after yoga (* indicates a significant difference (p < 0.05)).

[0055] The benefits of yoga have been studied in various fields, from chronic health conditions to mental disorders, and have been shown to be useful in improving overall health. In particular, yoga has been shown to be able to improve autonomic nervous function. Heart rate variability (HRV) at rest is typically used as a non-invasive measure of the autonomic regulation of heart rate. Alternatively, pulse rate variability (PRV) has been proposed as a substitute for HRV. This study aims to evaluate the effect of yoga on the autonomic nervous system by analyzing PRV obtained from NIVA signals. Before and after a yoga session by 20 healthy volunteers, time (statistics of normal-normal intervals), spectral (power in low and high frequency bands), and non-linear (delayed Poincaré plot analysis) parameters are analyzed. PRV analysis shows an increase in parameters related to parasympathetic activity and overall variability, as well as a decrease in parameters related to sympathetic activity and mean heart rate. These results support the beneficial effect of yoga on the autonomic nervous system in enhancing parasympathetic activity.

[0056] In some studies, peripheral venous lines of inpatients have been utilized and mainly used to directly administer fluids and drugs into the circulatory system to enable derivation of the patient's hemodynamic information. For example, continuous monitoring of the vascular volume status was used to detect bleeding at the initial stage 1, but no differences were seen in other physiological parameters such as heart rate, SpO2, or mean arterial pressure. Also, peripheral venous signals can provide additional information such as respiratory rate and pulse rate in addition to the patient's volume status.

[0057] This function of providing valuable additional information increases the interest in obtaining this signal in a non-invasive way that enables continuous monitoring in daily life. Non-invasive venous waveform analysis (NIVA) can acquire information regarding volume status, heart rate, and respiratory rate. The results show that the high-frequency power is significantly higher in PRV analysis from NIVA than in heart rate variability (HRV) from ECG, suggesting that NIVA signals can enhance the measurement of parasympathetic activity.

[0058] Some authors have studied the effect of yoga on autonomic function, showing an increase in parasympathetic activity through HRV analysis. We obtained both time-domain PRV parameters and spectral PRV parameters using a NIVA band before and after a yoga session by 20 volunteers. Furthermore, we propose a delay Poincaré plot (LPP) for non-linear analysis of the PRV series. When used together with time and frequency analysis of heart rate variability to study the autonomic regulation of the yoga and control groups of subjects, significant differences have already been found in the parameters of the Poincaré plot. Furthermore, the LPP has been shown to be reliable in ultra-short-term HRV analysis.

[0059] In this study, there were 5 men and 15 women with an age range of (27 - 79 years old, average 52 years old, median 56 years old). The subjects lay supine in the savasana position, supported by blankets under the head and lower limbs. Venous waveform signals were collected for 5 - 6 minutes before (pre-yoga) and after (post-yoga) a 2-hour "awareness yoga" session, which is a type of relaxation yoga with a slow recovery type. The venous waveforms were collected using a piezoelectric sensor placed on the palmar side of the wrist over the venous plexus, fixed with an elastic wrap. The sensor was interfaced with a prototype NIVA device (including an amplifier, a microcontroller, and a flash memory). The data was transferred to a computer as a text file using a USB cable and analyzed as follows.

[0060] The NIVA signal is first low-pass filtered at a cut-off frequency of 5 Hz to remove high-frequency noise. Frequencies below 0.3 Hz are also removed using a high-pass filter to remove low vibrations such as baseline fluctuations. In pulse detection, a linear-phase FIR low-pass differentiator (LPD) filter is first applied to emphasize the rising slope of the pulse, and then an adaptive threshold is applied to detect the peak of the LPD-filtered signal representing the point with the maximum slope of the NIVA signal. These k-th pulse detections are represented as n k as.

[0061] n kFrom this, the pulse interval series is d(k) = n k -n k-1 and is obtained. Pulses of the NIVA signal may be masked by noise, resulting in incorrect or missing pulse detections, and as a result, outlier (or abnormal value) pulse intervals occur. These outliers are identified and removed, and normal-normal (NN) intervals are obtained. FIG. 13 shows an example of pulse detection with incidence and its correction (upper panel), which is also reflected in the NN interval series (lower panel).

[0062] In the case of time-domain analysis, the PRV parameters are derived from the NN intervals. Three parameters, the standard deviation of the pulse intervals (SDNN), the square root of the mean squared difference of consecutive pulse intervals (RMSSD), and the mean pulse interval time [Number] are considered.

[0063] In the case of frequency-domain analysis, the PRV parameters are derived from the modulation signal m(n) which is assumed to have information regarding ANS activity, and are calculated and explained below. The instantaneous heart rate signal, d PR (n) is derived from d(k) according to a method based on a time-varying integral pulse frequency modulation (TVIPFM) model and resampled at 4 Hz. This signal is high-pass filtered to remove the average heart rate trend d PRM (n) (ultra-low frequency component), and is corrected to obtain the heart rate modulation signal m(n) = (d PR (n) - d PRM (n)) / d PRM (n).

[0064] In this method, the possibility of the existence of ectopic beats, false detections, or detection omissions is considered. However, if there are too many outliers relative to each other, the gap is too long, and artificial vibrations may be introduced by the estimated m(n). Therefore, if these gaps are longer than 3 seconds, usually these gaps are not considered. Only segments of 90 seconds or more without gaps exceeding 3 seconds were considered for reliable frequency domain analysis.

[0065] The power spectral density (PSD) of m(n) is calculated using the Welch periodogram with a 10-second overlap of a 60-second window. The power in the LF and HF bands was calculated by integrating the power spectra of the corresponding bands, PLF from 0.04 to 0.15 Hz and PHF from 0.15 to 0.4 Hz. The normalized LF power is P LFn =P LF / (P LF +P HF ) and is obtained as such.

[0066] The Poincaré plot is a graphical representation of the heartbeat dynamics triggered by the return map theory to explain the trajectory in phase space. In the standard version, the Poincaré plot is a scatter plot where each pulse interval d(k) is plotted against the previous pulse interval d(k - 1). In the delayed Poincaré plot (LPP) technique, a lag l is introduced, and a scatter plot is created by the points of coordinates d(k) and d(k - l). In previous studies, a range of lag values equal to 1 ≤ l ≤ 10 was investigated.

[0067] The most commonly used quantitative approach to describe the shape of the LPP is the elliptical fitting technique. According to this method, the LPP is rotated 45° clockwise, and the two standard deviations of the points around the vertical (SD1) axis and the horizontal (SD2) axis are calculated. SD1 is a measure of the short-term variability of the pulse interval series, while SD2 represents the long-term dynamics. In this study, the two standard deviations of SD1 and SD2 were investigated together with their ratio SD12 = SD1 / SD2.

[0068] The values of SD1, SD2, and SD12 were calculated in 50% overlapping windows lasting 35 seconds for both the pre-yoga and post-yoga NN intervals. In the LPP analysis, the standard range of lag values 1 ≤ l ≤ 10 was considered. The reliability of the LPP parameters calculated using the elliptical fitting technique in 35-second windows has already been investigated in previous studies through synthetic and actual data.

[0069] Considering each LPP parameter (SD1, SD2, and SD12) and each lag l, the median values between 35-second windows were calculated before and after the yoga session. For each subject, two median values were obtained for each LPP parameter and each lag corresponding to before and after yoga, respectively.

[0070] A paired Wilcoxon non-parametric statistical test was applied to all PRV parameters to examine the differences before and after the yoga session. Fifteen subjects were included in the analysis of time (SDNN, RMSSD, NN) and LLP parameters (SD1, SD2, SD12 for each lag), while nine subjects were included in the analysis of spectral parameters (P LFn and P HF ). When p < 0.05, the differences can be considered significantly different from zero.

[0071] Figure 14 shows an example of the NN interval series before and after the yoga session, and the corresponding PSD. After the yoga session, an overall increase in power is observed.

[0072] The time PRV parameters are shown in Figure 15. NN (corresponding to a decrease in average heart rate), SDNN, and RMSSD increased, but only SDNN was statistically significant. The spectral PRV parameters are shown in Figure 16. P LFn decreased and P HF increased, but in this case, neither was significant. The only LPP parameter found to be statistically different is the ratio SD12 (lags 5 and 7) shown in Figure 16, which shows a significant increase after yoga.

[0073] In some efforts, the role of yoga in overall health and a healthy lifestyle has been investigated. Yoga has also been proposed for the treatment of severe mental disorders. Significant benefits have been seen in reducing the general psychopathological assessment of schizophrenic patients, improving cognition and function, and reducing the severity of depressive symptoms. In these studies focusing on autonomic function, an increase in parasympathetic activity, a decrease in sympathetic activity, and an increase in overall heart rate variability were observed.

[0074] The inventors' results are consistent with those found in the literature. Most of the time and spectral parameters did not show significant differences, probably due to the small size of the data, but actually followed the same trend as other studies. An increase in NN (decrease in average heart rate) was also observed as in previous studies. In most recordings, a trend similar to the example shown in Figure 14 was seen. Before yoga, the heart rate decreased over time as the yoga session approached, while after yoga, the heart rate increased over time. The increase in SDNN is related to the overall increase in overall variability, while the increase in RMSSD is related to the increase in parasympathetic activity.

[0075] Spectral parameters also show similar results. There is a decrease in LF power (in normalized units), an increase in HF power, and a decrease in the LF / HF ratio, which corresponds to an increase in parasympathetic activity and a decrease in both sympathetic activity and sympathetic tone. Figure 14 also shows an increase in the power of both LF and HF, which may be related to an increase in parasympathetic activity affecting both bands. Using the function of the NIVA band to estimate the respiratory rate, it was found that the respiratory rate was within the conventional HF band (0.15 - 0.4 Hz) in all recordings, and there was no statistical difference in PHF when the HF band was concentrated on the respiratory rate compared to using the conventional HF band. The main limitation of spectral analysis is that the NIVA signal is very sensitive to movement artifacts, and for only 9 subjects, segments longer than 90 seconds without outliers in the pulse interval series were found. Also, the statistical test of the time parameters was repeated for only these 9 subjects, and the trend was the same, and at this time SDNN was not significant.

[0076] To overcome this limitation, other non-linear parameters that can be calculated with a very short time window, such as the delayed Poincaré plot, were used. Some previous studies reported a correlation between delayed SD1 and HF power and revealed its strong ability to detect an increase in vagal modulation. According to recent literature, the SD12 parameter is related to the non-linear component of heart rate dynamics (especially when l = 5, 6), and it is known that an increase in sympathetic activity decreases its value. Therefore, the increase in SD12 found in this study suggests a decrease in sympathetic activity. Also, when this analysis was repeated using the entire recording instead of a 35-second window, similar statistical differences were found in SD12 at lag 4 and lag 5.

[0077] In summary, it was shown that the NIVA band can detect changes in autonomic function after a yoga session. Similar to other studies on heart rate variability, yoga was found to increase parasympathetic activity and decrease sympathetic balance and mean heart rate. Overall variability also increased overall. To overcome the limitations seen in the NIVA signal that make it difficult to analyze PRV parameters in the frequency domain, time parameters and non-linear parameters were used and similar results to spectral parameters were found.

[0078] Although various exemplary aspects and exemplary embodiments are disclosed herein, other aspects and embodiments will be apparent to those of ordinary skill in the art. The various exemplary aspects and exemplary embodiments disclosed herein are for purposes of illustration and are not intended to be limiting, and the true scope and spirit thereof are indicated by the following claims.

Claims

1. 1. A method comprising: (a) generating, via a sensor of a computing device, a signal representative of vibrations originating from a blood vessel of a subject, the vibrations indicative of a heartbeat and / or respiration of the subject; (b) using the signal to determine a first time elapsed between each pair of successive heart beats indicated by the vibrations, and / or a second time elapsed between each pair of successive breaths indicated by the vibrations; (c) using the determined first elapsed time to determine heart rate variability of the subject, and / or using the determined second elapsed time to determine respiratory rate variability of the subject.

2. 10. The method of claim 1, wherein the sensor comprises a piezoelectric sensor, a pressure sensor, a capacitance sensor, a strain sensor, a force sensor, an optical wavelength selective reflectance or absorbance measurement system, a tonometer, an ultrasound probe, a plethysmograph, or a pressure transducer.

3. 3. The method of claim 1, wherein the vibrations comprise vibrations of the walls of the blood vessels produced by fluid flowing through the blood vessels, produced by wall tension of the blood vessels, or produced by contracting or relaxing the blood vessels in response to the fluid flowing through the blood vessels or by breathing of the subject.

4. 4. The method of claim 1, wherein the sensor is positioned in proximity to a peripheral vein or artery of the subject while the signal is being generated, and the vibrations originate from the peripheral vein or artery of the subject.

5. The method according to any one of claims 1 to 4, wherein the subject is a human subject or an animal subject.

6. 6. The method of claim 1, wherein using the signal to determine the first elapsed time comprises identifying successive heart rate peaks in the signal and determining the elapsed time between the successive heart rate peaks.

7. 7. The method of claim 1, wherein using the signal to determine the second elapsed time comprises identifying successive respiratory peaks in the signal and determining an elapsed time between the successive respiratory peaks.

8. The method of claim 6 or 7, wherein identifying the successive respiration or heart rate peaks of the signal comprises identifying successive peaks using an adaptive threshold.

9. 9. The method of claim 1, wherein using the signal to determine the first elapsed time and / or the second elapsed time comprises identifying consecutive times at which the signal has a maximum in a slope, and determining the elapsed time between the consecutive times at which the signal has a maximum in a slope.

10. The method of any preceding claim, wherein using the signal to determine the first elapsed time and / or the second elapsed time comprises filtering the signal with a low pass filter.

11. The method of claim 10 , wherein using the signal to determine the first elapsed time and / or the second elapsed time further comprises amplifying the filtered signal.

12. 12. The method of claim 11, wherein using the signal to determine the first elapsed time and / or the second elapsed time further comprises filtering the amplified signal with a low pass differentiator filter.

13. 13. The method of any of claims 1 to 12, wherein determining heart rate variability of the subject using the determined first elapsed time comprises calculating a variance of the first elapsed time.

14. 14. The method of any of claims 1 to 13, wherein determining respiration rate variability of the subject using the determined second elapsed time comprises calculating a variance of the second elapsed time.

15. The method of any of claims 1 to 14, wherein the first elapsed time corresponds to heart beats occurring within a duration in the range of 1 second to 10 seconds, 2 seconds to 8 seconds, or 2.5 seconds to 6 seconds.

16. 16. The method of claim 1, wherein the second elapsed time corresponds to a breath occurring within a duration in a range of 1 second to 10 seconds, 2 seconds to 8 seconds, or 2.5 seconds to 6 seconds.

17. 17. The method of any of claims 1-16, further comprising using the signal to determine a respiratory rate of the subject, wherein using the signal to determine the first elapsed time comprises using the signal to determine a time elapsed between each pair of successive heart beats occurring within a duration representing the inverse of the determined respiratory rate.

18. Determining the respiratory rate of the subject using the signal includes: obtaining a power spectral density of the signal; 20. The method of claim 17, comprising determining the respiration rate by: determining the lowest frequency of the power spectral density at which there is a peak that exceeds a predetermined power threshold.

19. The method of any of claims 1 to 18, further comprising assessing a parasympathetic nervous system response based on the determined heart rate variability and / or the respiratory rate variability.

20. 20. The method of claim 19, further comprising processing the signal into a frequency domain signal representing intensity with respect to frequency, and wherein assessing the parasympathetic nervous system response comprises assessing the parasympathetic nervous system response further based on a magnitude of a non-harmonic peak in the frequency domain signal corresponding to the respiratory rate.

21. The method of any of claims 1 to 20, further comprising assessing a response of the sympathetic nervous system based on the determined heart rate variability and / or the respiratory rate variability.

22. 22. The method of claim 21, further comprising processing the signal into a frequency domain signal representing intensity with respect to frequency, and wherein assessing the sympathetic nervous system response comprises assessing the sympathetic nervous system response further based on a magnitude of a non-harmonic peak in the frequency domain signal corresponding to the respiratory rate.

23. 23. The method of any of claims 1 to 22, wherein the method comprises performing steps (a) to (c) (i) before administering a treatment to the subject and (ii) after administering the treatment.

24. 24. The method of any of claims 1 to 23, further comprising using the determined heart rate variability and / or respiratory rate variability to treat or diagnose overtraining of athletes, optimization of athletic performance, post traumatic stress disorder, anxiety, pain, eating disorders, obesity, bariatric surgery, autonomic dysfunction, depression, opiate, alcohol or other substance addiction, delirium, acute or chronic illness, critical illness, post-operative stress, traumatic brain injury, cancer, post-myocardial infarction, complications caused by regional anesthesia or conscious sedation, burns, and / or complications of childbirth.

25. The method of any of claims 1 to 24, further comprising using the determined heart rate variability and / or the determined respiratory rate variability to determine the effectiveness of a treatment protocol.

26. 26. The method of claim 25, wherein the treatment protocol comprises oxytocin administration, narcotic administration, sedative administration, anxiolytic administration, behavioral therapy intervention, anxiolytic administration, behavioral therapy intervention, Reiki, healing touch, meditation, yoga, exercise, sympathetic neurolytic administration, alpha and beta antagonist administration, environmental intervention, music therapy, art therapy, virtual reality therapy, or pet therapy.

27. The method according to any one of claims 1 to 26, wherein the blood vessel is a vein.

28. The method according to any one of claims 1 to 26, wherein the blood vessel is an artery.

29. 1. A computing device comprising: one or more processors; A sensor; A computer readable medium storing instructions which, when executed by said one or more processors, cause said computing device to perform any of the methods according to claims 1 to 28.

30. 30. The computing device of claim 29, comprising: a master device housing the one or more processors and the computer readable medium; and a slave device housing the sensor.

31. The computing device of claim 30 , wherein the master device is a wearable device.

32. The computing device of claim 30 , wherein the master device is a mobile phone.

33. The computing device of claim 30 , wherein the master device is the tablet computer.

34. The computing device of claim 30 , wherein the master device is a laptop computer.

35. The computing device of claim 30 , wherein the master device is a desktop computer.

36. A non-transitory computer readable medium storing instructions that, when executed by any of the computing devices of claims 29 to 35, cause said computing device to perform any of the methods of claims 1 to 28.

Images