System and method for determining blood flow using multi-wavelength photoplethysmography

Abstract

A system and method for determining blood flow using multi-wavelength photoplethysmography is provided. The method including: receiving the photoplethysmography signals from an optical input sensor positioned to capture the photoplethysmography signals from the skin of a user, the photoplethysmography signals including at least two different optical wavelengths; determining arteriolar pulse transit time using time delays determined between the photoplethysmography signals of different wavelengths; determining skin blood flow as a function of the determined arteriolar pulse transit time; and outputting the skin blood flow.

Description

SYSTEM AND METHOD FOR DETERMINING BLOOD FLOW USING MULTI-WAVELENGTH PHOTOPLETHYSMOGRAPHYTECHNICAL FIELD

[0001] The following relates generally to photoplethysmography; and more specifically, to a system and method for determining blood flow using multi-wavelength photoplethysmography.BACKGROUND

[0002] Cardiovascular diseases (CVD) are estimated to be responsible for 31% of all worldwide deaths. The skin offers an accessible vascular bed for assessing endothelial microvascular function in humans through skin blood flow measurements. With microvascular dysfunction often appearing early in disease progression, its detection can offer early prognostic information for disease prevention. Continuous and remote monitoring of skin blood flow can provide key information on progression of various disease types, and can serve as a cardiovascular risk stratification tool. Peripheral microvascular function assessment through blood flow generally reflects coronary microvascular status.SUMMARY

[0003] In an aspect, there is provided a method for determining blood flow using multiwavelength photoplethysmography signals, the method executed on one or more processing units in communication with a data storage, the method comprising: receiving the photoplethysmography signals from an optical input sensor positioned to capture the photoplethysmography signals from the skin of a user, the photoplethysmography signals comprising at least two different optical wavelengths; determining arteriolar pulse transit time using time delays determined between the photoplethysmography signals of different wavelengths; determining skin blood flow as a function of the determined arteriolar pulse transit time; and outputting the skin blood flow.

[0004] In a particular case of the method, the at least two different optical wavelengths comprise a first wavelength below 600 nanometers and a second wavelength above 600 nanometers.

[0005] In another case of the method, determining the arteriolar pulse transit time comprises using a reference wavelength and determining time delays for other wavelengths relative to the reference wavelength.

[0006] In yet another case of the method, the reference wavelength is the deepest-penetrating wavelength.

[0007] In yet another case of the method, determining the arteriolar pulse transit time comprises determining a cross-correlation between the photoplethysmography signals of different wavelengths with a sliding window.

[0008] In yet another case of the method, determining the arteriolar pulse transit time comprises identifying a time delay corresponding to a maximum of the cross-correlation.

[0009] In yet another case of the method, the method further comprising interpolating the photoplethysmography signals to a common time vector using spline interpolation.

[0010] In yet another case of the method, the method further comprising bandpass filtering the photoplethysmography signals to isolate pulsatile human heartrate.

[0011] In yet another case of the method, the arteriolar pulse transit time is determined as a linear inverse relationship with the time delays.

[0012] In yet another case of the method, the arteriolar pulse transit time is determined using higher order polynomials of the time delays.

[0013] In another aspect, there is provided a system for determining blood flow using multiwavelength photoplethysmography signals, the system comprising one or more processing units in communication with a data storage, the data storage comprising instructions for the one or more processing units to execute: an input module to receive the photoplethysmography signals from an optical input sensor positioned to capture the photoplethysmography signals from the skin of a user, the photoplethysmography signals comprising at least two different optical wavelengths; a time-delay module to determine an arteriolar pulse transit time using time delays determined between the photoplethysmography signals of different wavelengths; a blood flow module to determine skin blood flow as a function of the determined arteriolar pulse transit time; and an output module to output the skin blood flow.

[0014] In a particular case of the system, the at least two different optical wavelengths comprise a first wavelength below 600 nanometers and a second wavelength above 600 nanometers.

[0015] In another case of the system, determining the arteriolar pulse transit time comprises using a reference wavelength and determining time delays for other wavelengths relative to the reference wavelength.

[0016] In yet another case of the system, the reference wavelength is the deepest-penetrating wavelength.

[0017] In yet another case of the system, determining the arteriolar pulse transit time comprises determining a cross-correlation between the photoplethysmography signals of different wavelengths with a sliding window.

[0018] In yet another case of the system, determining the arteriolar pulse transit time comprises identifying a time delay corresponding to a maximum of the cross-correlation.

[0019] In yet another case of the system, the one or more processing units further executing a filtering module to interpolate the photoplethysmography signals to a common time vector using spline interpolation.

[0020] In yet another case of the system, the one or more processing units further executing a filtering module to bandpass filter the photoplethysmography signals to isolate pulsatile human heartrate.

[0021] In yet another case of the system, the arteriolar pulse transit time is determined as a linear inverse relationship with the time delays.

[0022] In another aspect, there is provided a non-transitory computer-readable medium storing instructions that, when executed by one or more processing units, cause the one or more processing units to perform operations comprising: receiving photoplethysmography signals from an optical input sensor positioned to capture the photoplethysmography signals from the skin of a user, the photoplethysmography signals comprising at least two different optical wavelengths; determining arteriolar pulse transit time using time delays determined between the photoplethysmography signals of different wavelengths; determining skin blood flow as a function of the determined arteriolar pulse transit time; and outputting the skin blood flow.

[0023] These and other aspects are contemplated and described herein. It will be appreciated that the foregoing summary sets out representative aspects of embodiments to assist skilled readers in understanding the following detailed description.BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The features of the invention will become more apparent in the following detailed description in which reference is made to the appended drawings wherein:

[0025] FIG. 1 shows a schematic diagram of a system for determining blood flow using multiwavelength photoplethysmography, in accordance with an embodiment;

[0026] FIG. 2 is a flowchart of a method determining blood flow using multi-wavelength photoplethysmography, in accordance with an embodiment;

[0027] FIG. 3 is a diagram showing simultaneous wavelengths and distance-based light paths through human skin with multi-wavelength photoplethysmography (PPG) and to determine arteriolar pulse transit time;

[0028] FIG. 4 illustrates a diagram of simultaneous data capture of both multi-wavelength PPG sensor and a Laser Doppler Flowmetry (LDF) sensor on opposing hands of the same person, in accordance with example experiments;

[0029] FIGS. 5A to 5C illustrate experimental results from multi-wavelength PPG sensor placed on a finger compared to LDF sensor;

[0030] FIG. 6 are illustrations showing a wearable sensor with a multiwavelength sensor and multiple pairs of individually controllable LEDs, in accordance with the system of FIG. 1 ;

[0031] FIG. 7 is a diagram illustrating an architecture of a microvascular bed with representative microfluidic dynamic in (a) normal and (b) vasoconstricting conditions;

[0032] FIG. 8 is a chart illustrating an example of spectrometer measurements of the six LEDs incorporated within the wearable sensor in the example experiments;

[0033] FIG. 9 is a diagram illustrating the simultaneous data collection of the multi-wavelength PPG sensor and the LDF sensor during baseline and a cold pressor, for the example experiments;

[0034] FIG. 10 is a plot showing an arterial pulse transit time metric measured using the system of FIG. 1 and the LDF measurements, plotted together for the data collected simultaneously during baseline and a cold pressor stimulus, in the example experiments;

[0035] FIG. 11 is a correlation plot showcasing the strength of the correlation between the measurements of the system of FIG. 1 and the LDF measurements, in the example experiments;

[0036] FIG. 12 illustrates a Bland-Altman plot for the cold pressor stimulus, presenting the level of agreement between the measurements of the system of FIG. 1 and the LDF measurements, in the example experiments;

[0037] FIG. 13 illustrates another Bland-Altman plot for the cold pressor stimulus, in further example experiments;

[0038] FIG. 14 is a plot showing an arterial pulse transit time metric measured using the system of FIG. 1 and the LDF measurements, plotted together for the data collected simultaneously during baseline and a breath hold stimulus and release, in further example experiments;

[0039] FIG. 15 illustrates a Bland-Altman plot for the breath hold stimulus and release, presenting the level of agreement between the measurements of the system of FIG. 1 and the LDF measurements, in the example experiments;

[0040] FIG. 16 is a plot showing an arterial pulse transit time metric measured using the system of FIG. 1 and the LDF measurements, plotted together for the data collected simultaneously during baseline, a finger occlusion stimulus, and post-release reactive hyperemia, in further experiments;

[0041] FIG. 17 illustrates a Bland-Altman plot for the finger occlusion stimulus, presenting the level of agreement between the measurements of the system of FIG. 1 and the LDF measurements, in the example experiments; and

[0042] FIG. 18 illustrates charts showing a raw PPG signal (top) and arterial pulse transit time (bottom) determined by the system of FIG. 1 for two wavelengths (530nm and 740nm), in the example experiments.DETAILED DESCRIPTION

[0043] Embodiments will now be described with reference to the figures. For simplicity and clarity of illustration, where considered appropriate, reference numerals may be repeated among the Figures to indicate corresponding or analogous elements. In addition, numerous specific details are set forth in order to provide a thorough understanding of the embodiments described herein. However, it will be understood by those of ordinary skill in the art that the embodiments described herein may be practiced without these specific details. In other instances, well-known methods, procedures and components have not been described in detail so as not to obscure the embodiments described herein. Also, the description is not to be considered as limiting the scope of the embodiments described herein.

[0044] Various terms used throughout the present description may be read and understood as follows, unless the context indicates otherwise: “or” as used throughout is inclusive, as though written “and / or”; singular articles and pronouns as used throughout include their plural forms, and vice versa; similarly, gendered pronouns include their counterpart pronouns so that pronouns should not be understood as limiting anything described herein to use, implementation, performance, etc. by a single gender; “exemplary” should be understood as“illustrative” or “exemplifying” and not necessarily as “preferred” over other embodiments. Further definitions for terms may be set out herein; these may apply to prior and subsequent instances of those terms, as will be understood from a reading of the present description.

[0045] Any module, unit, component, server, computer, terminal, engine or device exemplified herein that executes instructions may include or otherwise have access to computer readable media such as storage media, computer storage media, or data storage devices (removable and / or non-removable) such as, for example, magnetic disks, optical disks, or tape. Computer storage media may include volatile and non-volatile, removable and non-removable media implemented in any method or technology for storage of information, such as computer readable instructions, data structures, program modules, or other data. Examples of computer storage media include RAM, ROM, EEPROM, flash memory or other memory technology, CD- ROM, digital versatile disks (DVD) or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can be accessed by an application, module, or both. Any such computer storage media may be part of the device or accessible or connectable thereto. Further, unless the context clearly indicates otherwise, any processor or controller set out herein may be implemented as a singular processor or as a plurality of processors. The plurality of processors may be arrayed or distributed, and any processing function referred to herein may be carried out by one or by a plurality of processors, even though a single processor may be exemplified. Any method, application or module herein described may be implemented using computer readable / executable instructions that may be stored or otherwise held by such computer readable media and executed by the one or more processors.

[0046] The following relates generally to photoplethysmography; and more specifically, to a system and method for remote monitoring of blood flow using multi-wavelength photoplethysmography.

[0047] There exist a number of approaches for microvascular blood flow measurement. For example, using Laser Doppler Flowmetry (LDF). In LDF, an optical fibre is used to illuminate the skin by a laser light. The light reflected back from the tissue is collected by another fibre(s). Moving particles, such as red blood cells, cause a Doppler shift that broadens the spectrum of the reflected light. Laser Doppler flowmeters produce an output signal that is proportional to the blood cell perfusion (or flux), which represents the transport of blood cells through the microvasculature. Microvascular perfusion is the product of the mean cell velocity and mean blood cell concentration present in the volume of tissue under illumination from the laser beam.Generally, the equipment required to perform LDF measurements is bulky and lacks sufficient scalability. Thus, LDF cannot be easily implemented into remote monitoring devices, such as wearable fitness trackers, for remote and continuous monitoring purposes.

[0048] In another example, ultrasound Doppler can be used as an imaging technique in medical diagnostics, particularly in the assessment of blood flow. This approach operates based on the Doppler effect, which is the change in frequency or wavelength of a wave in relation to an observer who is moving relative to the wave source. In the case of ultrasound, it involves the reflection of high-frequency sound waves off moving blood cells. Ultrasound dopplers can provide information on directionality, velocity and volume flow of blood flow, as well as detecting abnormalities or obstruction that may impede blood flow. Generally, however, ultrasound Doppler is very operator-dependent, required highly skilled individuals to perform the procedure. Even among highly skilled operators, there can be great variability in measurements as Doppler measurements can be angle-dependent. Ultrasound Doppler is generally restricted to larger arteries and veins due to its performance degrading significantly when assessing blood flow in smaller arterioles and capillary beds. Further, ultrasound Doppler is generally costly and cannot be implemented in other form factors.

[0049] In another example, Laser Speckle Imaging (LSI) is an imaging technique used to visualize blood flow dynamics in tissues. It relies on the phenomenon of speckle contrast, which arises when coherent light, such as that from a laser, illuminates a rough surface or scatters off moving particles, such as red blood cells in blood vessels. LSI generally has lower spatial resolution that limits its ability to resolve small vessels or distinguish subtle changes in blood flow. Additionally, the LSI is generally expensive to purchase and maintain and the complexity of imaging set up and data analysis requires specialized training and expertise.

[0050] In contrast to the above approaches, the present embodiments are operatorindependent, low cost, low complexity, and provide approaches that are easily implementable in various form factors of wearable devices for optimal scalability and accessibility. In an example, the present embodiments can advantageously be implemented in consumer-grade fitness trackers. The present embodiments can provide a particularly useful biometric parameter for health monitoring, enabling more advanced remote monitoring and health tracking capabilities to assess risk, and intervene before devastating clinical events such as peripheral artery disease, heart failure, and other cardiovascular diseases. In other examples, the present embodiments can be used for post-surgery monitoring, wound healing assessment, and prescribed treatment validation.

[0051] Advantageously, the present embodiments are able to measure arterial (or arteriolar) pulse transit time (aPTT); which quantifies a pressure wave of blood as it propagates through skin tissue. aPTT refers to the time the pulse wave travels from deeper arteries, vertically through the arterioles, and to the microvascular capillary beds. In this way, the present embodiments measure the transit time of the pressure wave of blood as it propagates from deep skin tissue to, for example, the capillary bed of the skin. In order to do so, the present embodiments employ different wavelengths of light that each penetrate to different depths beneath the skin.

[0052] Referring now to FIG. 1, a system 50 for determining blood flow using multi-wavelength photoplethysmography, in accordance with an embodiment, is shown. The system 50 can be run on any suitable computing device, for example, on a general-purpose computing device, on a wearable device (such as a fitness tracker or a smartwatch), on a purpose-built controller, on remotely located servers (such as in a cloud-based arrangement), or the like. In some embodiments, the components of the system 50 are stored by and executed on a single computer system or device. In other embodiments, the components of the system 50 are distributed among two or more computer systems or devices that may be locally or remotely distributed.

[0053] FIG. 1 shows various physical and logical components of an embodiment of the system 50. As shown, the system 50 has a number of physical and logical components, including a processing unit 52, a data storage 54, a user interface 56, a device interface 60 and a local bus 80 enabling the processing unit 52 to communicate with the other components. The processing unit 102 executes various conceptual modules, as described herein in greater detail. The data storage 54 provides responsive data storage to the processing unit 52, including computerexecutable instructions for implementing the modules, as well as any data used by these services. The data storage 54 can include a non-transitory computer-readable medium. The user interface 106 enables a user to provide input via an input device, for example a touchscreen or button(s). The user interface 106 can also output information to output devices to the user, such as a display and / or speakers. The device interface 110 permits communication with various devices, such as a photoplethysmography optical input sensor.

[0054] In an embodiment, the processing unit 52 can execute a number of conceptual modules, which can include a device module 70, a simulation module 72, and a control module 74. In some cases, the functions and / or operations of the conceptual modules can be combined or executed on other modules.

[0055] Multi-wavelength Photoplethysmography (MW-PPG) sensors can include an optical light source (such as LEDs with different wavelengths) that are used to illuminate the skin of a user and also include photodetectors that measure changes in the absorption and reflection of light as it passes through the tissue. The system 50 can be used to provide critical vascular data in the form of arteriolar pulse transit time (aPTT). aPTT represents the time a pulse wave travels from deeper arteries, vertically through the arterioles, and to the superficial capillary beds; which is illustrated in the example diagram of FIG. 3. FIG. 3 is a diagram showing simultaneous wavelengths and distance-based light paths through human skin with multi-wavelength PPG and to determine arteriolar pulse transit time; which is referred to herein as ‘time delay’.

[0056] MW-PPG sensors that capture a wide spectrum of different wavelengths permit pulsatile waveforms from different skin depths to be captured; providing valuable information on pulse wave propagation through the tissue. Multi-wavelength PPG can be used as an indicator of systemic vascular resistance (SVR), and mean blood pressure (MBP) because SVR is a determinant thereof. However, the use of aPTT to provide physiological information on blood flow has been determined by the present inventors to provide useful results, and thus, is particularly advantageous.

[0057] FIG. 2 illustrates a flowchart of a method 200 for determining blood flow using multiwavelength photoplethysmography, in accordance with an embodiment.

[0058] At block 202, the input module 70 receives photoplethysmography (PPG) signals from an optical input sensor via the device interface 110. The optical input sensor is positioned to receive signals from the skin of a user. The PPG signals are received from a minimum of two different optical wavelengths. In a particular case, the optical input sensor can be comprised of a plurality of photodiodes. In a particular case, a first wavelength provides shallow penetration of the skin, such as at a wavelength below 600 nanometers (nm), and a second wavelength provides deeper penetration of the skin, such as at a wavelength above 600 nm. However, it is understood that any suitable wavelengths and photodiode spacings can be used, or combinations thereof. In further cases, additional wavelengths can also be received. In some cases, the input module 70 can interpolate the PPG data from the various wavelengths to a common time vector using, for example, spline interpolation.

[0059] At block 204, in most cases, the filtering module 72 performs filtering on the received PPG signals. The filtering can include bandpass filtering of the PPG signals to filter out noise; such as low-pass and high-pass Butterworth filters. Generally, the filtering includes a window size that is set such that the PPG signal isolates pulsatile human heartrate. In further cases, itshould be understood that other filters can be used, such as Chebychev and Savitzky-golay filters. Various cut off values can be used to isolate for human heartrate, for example, between 0.5Hz (low pass) and 5Hz (high pass).

[0060] At block 206, the time-delay module 74 determines a sliding window cross-correlation between particular segments of the filtered data and from this, determines time delays (TD). The sliding window can have any suitable definition, for example, between one pulse and 30 seconds. The defined window slides for a determined period and the cross-correlation determination can be repeated. In an example, the sliding period can be one second, but any suitable value can be used. In some cases, the sliding window can act as a data buffer such that as more filtered data is received, it is added to such buffer and older data is removed. In some cases, the time delays can be stored in a matrix data structure in order to permit structured data management and analysis when dealing with multiple signals, conditions, or experimental configurations; however any suitable data structure can be used. The time delays represent delays between the two or more wavelengths of the received and filtered PPG signal. In further cases, other approaches for determining time delays can be used, such as using findpeaks, using Principal Component Analysis-based signal decomposition, and using Hilbert's transforms.

[0061] At block 208, the blood flow module 76 determines an arteriolar pulse transit time (aPTT) as a cross-correlation between the determined time delays. The aPTT can be determined based on the time delay between two different wavelengths at a particular time. At block 210, the blood flow module 76 determines skin blood flow (SkBF) using a correlation to the aPTT determination. In many cases, the determination of the blood flow signal can use a metric determined for a particular body part. In an example, blood flow (e.g., in blood perfusion units) can be determined using a linear inverse relationship with the determined time delay. In some cases, higher order polynomials can be used for such determination to explain the relationship between the SkBF and the time delay (TD); for example, through fitting a polynomial model to capture the relationship. In a further example, blood flow can be determined using principle component analysis to reduce a plurality of wavelengths to two, such that aPTT can be determined from the cross-correlation. However, it should be understood that any suitable metric can be used to determine the blood flow signal from the determined time delays. At block 212, the output module 78 outputs the determined blood flow signal and / or skin haemodynamics to the data storage 54, the user interface 56, and / or the device interface 60.

[0062] In essence, cross-correlation determines a phase or time delay between two signals. The time delay physiologically relates to the propagation time of the blood pressure wave as the wave propagates from deep tissue / arteries to shallow tissue / capillaries. A sliding window advantageously facilitates determination of aPTT on smaller segments of data at a time; for example, of segments of less than 30 seconds. The cross-correlation generally outputs one value, and in this way, if aPTT actively changes, such as with changes in blood flow, the system 50 only needs an adequately small window to observe changes in aPTT. In most cases, a one pulse onset is used to determine aPTT and a larger window, encompassing multiple pulses, can potentially be more resistant to noise due to motion and other artifacts. For each window segment, cross-correlation can be determined between the signals within such window by extracting a segment of the signal for each window position. The cross-correlation function will generally produce a set of values for different delay periods. The delay at maximum correlation corresponds to the aPTT (and thus, the time delay).

[0063] In many cases, the time delay is determined between two wavelengths at a time by using a reference wavelength (for example, the deepest penetrating wavelength), and then determining a time delay for other wavelengths individually in comparison to this reference wavelength. Each comparison generates an aPTT that originates from different depths within the skin tissue. In this way, the system 50 can quantify not just deep-to-shallow pulse transit times, but also deep-to-medium, and medium-to-shallow times; and the like.

[0064] To validate the present embodiments’ ability to determine accurate measures of blood flow, the present inventors conducted example experiments. In the example experiments, data was collected using the multi-wavelength photoplethysmography sensor (MW-PPG) and compared to data collected using a laser doppler flowmetry device. To establish a correlation between MW-PPG derived arteriolar pulse transit time and Laser Doppler Flowmetry SkBF, data was collected using both the MW-PPG sensor and Laser doppler flowmetry device simultaneously. FIG. 4 illustrates a diagram of simultaneous data capture of both MWPPG sensor and LDF on opposing hands of the same person.

[0065] In the example experiments, to determine correlation between the MW-PPG aPTT and the LDF SkBF, the aPTT determinations were plotted on top of the LDF for correlation assessment. The aPTT data was inverted on the y-axis. When plotted together, it was observed that there was a strong agreement between the inverted aPTT plot and SkBF measurements plotted, as illustrated in the plots of FIGS. 5A to 5C. This held true across multiple data capture sessions within an individual and across different individuals as well.

[0066] In an example, the relationship can be determined per the following correlation, however, any suitable correlation can be used:Skin Blood flow (SkBF) = - - “ - (1)Where a is a scaling constant that determines the maximum possible value of the SkBF when arteriolar PTT is zero; permitting the setting of an upper bound if ft is very small compared to the arteriolar PTT. Where ft is a baseline offset that shifts the time arteriolar PTT in the denominator; ensuring that the value is not undefined or excessively large when the arteriolar PTT is very small or zero. Both a and ft can be manually adjusted based on fit.

[0067] FIGS. 5A to 50 illustrate experimental results from MW-PPG sensor placed on a finger compared to LDF. SkBF measurements were captured at baseline across two individuals (1 male, 1 female) over different sessions. As shown in the outputted charts, comparisons of MW- PPG derived aPTT to skin blood flow measurements captured by a Laser Doppler Flowmetry showed a strong agreement between the inverse of the aPTT metric and the Laser Doppler SkBF measurements; illustrating the accuracy of the aPTT measurement determined by the present embodiments.

[0068] In another example experiment, the system 50 was incorporated into a wearable device and included six light-emitting-diodes (LED), each emitting a separate wavelength, and four concentric photodiodes that can quantify pulse wave dynamics within the skin of the user. This example embodiment is illustrated in FIG. 6, which shows the wearable sensor with a multiwavelength sensor (i.e. , photodiodes (PD)) and multiple pairs of individually controllable LEDs. In this example, there are four PDs highlighted in exterior boxes and six LEDs located in the center of the PDs.

[0069] Using these multiple LEDs emitting different wavelengths, longer wavelengths of emitted light can be used to penetrate deeper into the skin, allowing signals to be picked up from the deeper arteries; while shorter wavelengths of light penetrate shallower capillary beds. Arterioles, very small vessels with diameter from 5-100 pm, play a vital role in maintaining homeostasis through regulation of microfluidic dynamics in the microvascular capillary beds; whereby the blood flow is regulated by active response to vasomodulatory stimulus. Under normal conditions, blood flows into the capillary bed from the arterioles, while during vasoconstricting conditions, blood flow is reduced in the capillary beds or may be shunted directly from the arterioles to the venules, bypassing the capillary bed. FIG. 7 is a diagram illustrating the architecture of a microvascular bed with representative microfluidic dynamic in (a) normal and(b) vasoconstricting conditions. FIG. 8 is a chart illustrating an example of spectrometer measurements of the six LEDs incorporated within the wearable sensor.

[0070] In the example experiments, data was collected simultaneously using the system 50 and LDF during baseline and a cold pressor event induced by a unilateral foot submersion into ice water; i.e., a vasoconstricting stimulus. FIG. 9 is a diagram illustrating the simultaneous data collection of the system 50 (referred to as a MWPPG sensor) and the LDF during baseline and a cold pressor, unilateral foot ice bath, to establish correlation to validate the aPTT metric as a blood flow metric.

[0071] In other experiments, two additional vasomodulatory stimuli were evaluated to further validate and evaluate the accuracy and robustness of the aPTT metric and its correlation to the reference LDF measurement. One of the additional stimuli was a breath hold stimulus in which a maximum inspiratory breath hold was maintained for one minute and then released (twice); where CO2 build up led to vasodilation upon release. The other additional stimulus was a local occlusion stimulus in which a five-minutes finger occlusion caused temporary ischemia followed by reactive hyperemia (vasodilation). Measurements from the multi-wavelength PPG sensor and LDF were captured simultaneously on fingers of contralateral hands.

[0072] FIG. 10 is a plot showing the aPTT metric measured using the system 50 and the LDF measurements plotted together for the data collected simultaneously; including showing where the cold pressor was applied. A correlation between the inverse of the aPTT metric determined by the system 50 to the referenced LDF measurements (SkBF) was established with a r2value of 0.74 and the relationship represented by Equation (1). FIG. 11 is a correlation plot showcasing the strength of the correlation between the two measurements. A Bland-Altman analysis was performed to measure agreement between the two approaches yielded a mean difference of 36 and the limits of agreement (LOA), calculated as ±1.96 standard deviations from the mean difference, ranged from -56 to 130. FIG. 12 illustrates a Bland-Altman plot presenting the level of agreement between the two measurements. An additional Bland-Altman plot for the cold pressor stimulus, obtained in further example experiments (r2value of 0.79), is shown in FIG. 13. Accordingly, the system 50 was able to quantify microvascular blood flow sufficiently accurately when compared to the much more expensive and significantly less portable LDF approach.

[0073] Comparisons of MW-PPG derived aPTT to SkBF measured by LDF demonstrated that the inverse of the aPTT metric tracked the LDF SkBF across the evaluated vasomodulatory stimuli. FIG. 14 shows an overlay of LDF SkBF and the inverse of the aPTT metric acrossbaseline and a breath hold stimulus with release (maximum inspiratory breath hold for one minute, performed twice). FIG. 15 shows a corresponding Bland-Altman plot presenting the level of agreement between the measurements (r2value of 0.63). FIG. 16 shows an overlay of LDF SkBF and the inverse of the aPTT metric across baseline, a local occlusion stimulus (five- minutes finger occlusion), and post-release reactive hyperemia. FIG. 17 shows a corresponding Bland-Altman plot presenting the level of agreement (r2value of 0.81) between the measurements.

[0074] Together with the Bland-Altman plot for the cold pressor stimulus (FIG. 12) and another Bland-Altman plot obtained in further example experiments (FIG. 13), these results illustrate the robustness of the aPTT-derived blood-flow metric under vasomodulatory conditions.

[0075] The results of the example expierments further validate the substantial advantages of using aPTT as a blood flow metric, showcasing the robustness of the aPTT metric for capturing blood flow changes that may be physiologically experienced under various conditions.

[0076] In accordance with further example experiments, FIG. 18 illustrates charts showing a raw PPG signal (top) and aPTT determined by the system 50 (bottom) for two wavelengths (530nm and 740nm).

[0077] Although the invention has been described with reference to certain specific embodiments, various modifications thereof will be apparent to those skilled in the art without departing from the spirit and scope of the invention as outlined in the claims appended hereto.

Claims

CLAIMS1. A method for determining blood flow using multi-wavelength photoplethysmography signals, the method executed on one or more processing units in communication with a data storage, the method comprising: receiving the photoplethysmography signals from an optical input sensor positioned to capture the photoplethysmography signals from the skin of a user, the photoplethysmography signals comprising at least two different optical wavelengths; determining arteriolar pulse transit time using time delays determined between the photoplethysmography signals of different wavelengths; determining skin blood flow as a function of the determined arteriolar pulse transit time; and outputting the skin blood flow.

2. The method of claim 1 , wherein the at least two different optical wavelengths comprise a first wavelength below 600 nanometers and a second wavelength above 600 nanometers.

3. The method of claim 1, wherein determining the arteriolar pulse transit time comprises using a reference wavelength and determining time delays for other wavelengths relative to the reference wavelength.

4. The method of claim 3, wherein the reference wavelength is the deepest-penetrating wavelength.

5. The method of claim 1, wherein determining the arteriolar pulse transit time comprises determining a cross-correlation between the photoplethysmography signals of different wavelengths with a sliding window.

6. The method of claim 5, wherein determining the arteriolar pulse transit time comprises identifying a time delay corresponding to a maximum of the cross-correlation.

7. The method of claim 1, further comprising interpolating the photoplethysmography signals to a common time vector using spline interpolation.

8. The method of claim 1, further comprising bandpass filtering the photoplethysmography signals to isolate pulsatile human heartrate.

9. The method of claim 1, wherein the arteriolar pulse transit time is determined as a linear inverse relationship with the time delays.

10. The method of claim 1 , wherein the arteriolar pulse transit time is determined using higher order polynomials of the time delays.

11. A system for determining blood flow using multi-wavelength photoplethysmography signals, the system comprising one or more processing units in communication with a data storage, the data storage comprising instructions for the one or more processing units to execute: an input module to receive the photoplethysmography signals from an optical input sensor positioned to capture the photoplethysmography signals from the skin of a user, the photoplethysmography signals comprising at least two different optical wavelengths; a time-delay module to determine an arteriolar pulse transit time using time delays determined between the photoplethysmography signals of different wavelengths; a blood flow module to determine skin blood flow as a function of the determined arteriolar pulse transit time; and an output module to output the skin blood flow.

12. The system of claim 11, wherein the at least two different optical wavelengths comprise a first wavelength below 600 nanometers and a second wavelength above 600 nanometers.

13. The system of claim 11, wherein determining the arteriolar pulse transit time comprises using a reference wavelength and determining time delays for other wavelengths relative to the reference wavelength.

14. The system of claim 13, wherein the reference wavelength is the deepest-penetrating wavelength.

15. The system of claim 11, wherein determining the arteriolar pulse transit time comprises determining a cross-correlation between the photoplethysmography signals of different wavelengths with a sliding window.

16. The system of claim 15, wherein determining the arteriolar pulse transit time comprises identifying a time delay corresponding to a maximum of the cross-correlation.

17. The system of claim 11, further comprising a filtering module to interpolate the photoplethysmography signals to a common time vector using spline interpolation.

18. The system of claim 11, further comprising a filtering module to bandpass filter the photoplethysmography signals to isolate pulsatile human heartrate.

19. The system of claim 11, wherein the arteriolar pulse transit time is determined as a linear inverse relationship with the time delays.

20. A non-transitory computer-readable medium storing instructions that, when executed by one or more processing units, cause the one or more processing units to perform operations comprising: receiving photoplethysmography signals from an optical input sensor positioned to capture the photoplethysmography signals from the skin of a user, the photoplethysmography signals comprising at least two different optical wavelengths; determining arteriolar pulse transit time using time delays determined between the photoplethysmography signals of different wavelengths; determining skin blood flow as a function of the determined arteriolar pulse transit time; and outputting the skin blood flow.

Images