IOP PUBLISHING
`
`Physiol. Meas. 28 (2007) R1–R39
`
`PHYSIOLOGICAL MEASUREMENT
`
`doi:10.1088/0967-3334/28/3/R01
`
`TOPICAL REVIEW
`
`Photoplethysmography and its application in clinical
`physiological measurement
`
`John Allen
`
`Regional Medical Physics Department, Freeman Hospital, Newcastle upon Tyne NE7 7DN, UK
`
`E-mail: john.allen@nuth.nhs.uk
`
`Received 8 October 2006, accepted for publication 24 January 2007
`Published 20 February 2007
`Online at stacks.iop.org/PM/28/R1
`
`Abstract
`Photoplethysmography (PPG) is a simple and low-cost optical technique that
`can be used to detect blood volume changes in the microvascular bed of tissue.
`It is often used non-invasively to make measurements at the skin surface. The
`PPG waveform comprises a pulsatile (‘AC’) physiological waveform attributed
`to cardiac synchronous changes in the blood volume with each heart beat,
`and is superimposed on a slowly varying (‘DC’) baseline with various lower
`frequency components attributed to respiration, sympathetic nervous system
`activity and thermoregulation. Although the origins of the components of
`the PPG signal are not fully understood, it is generally accepted that they
`can provide valuable information about the cardiovascular system. There has
`been a resurgence of interest in the technique in recent years, driven by the
`demand for low cost, simple and portable technology for the primary care
`and community based clinical settings, the wide availability of low cost and
`small semiconductor components, and the advancement of computer-based
`pulse wave analysis techniques. The PPG technology has been used in a
`wide range of commercially available medical devices for measuring oxygen
`saturation, blood pressure and cardiac output, assessing autonomic function
`and also detecting peripheral vascular disease. The introductory sections of the
`topical review describe the basic principle of operation and interaction of light
`with tissue, early and recent history of PPG, instrumentation, measurement
`protocol, and pulse wave analysis. The review then focuses on the applications
`of PPG in clinical physiological measurements, including clinical physiological
`monitoring, vascular assessment and autonomic function.
`
`ageing, artery, autonomic function, blood pressure, cardiac
`Keywords:
`output, cardiovascular, diabetes, endothelial function, heart rate, infrared,
`microcirculation, photoplethysmography (PPG), pulse wave
`analysis,
`Raynaud’s phenomenon, vascular disease, vein
`
`0967-3334/07/030001+39$30.00 © 2007 IOP Publishing Ltd Printed in the UK
`
`R1
`
`Apple Inc.
`APL1021
`U.S. Patent No. 8,929,965
`
`FITBIT, Ex. 1021
`
`

`

`R2
`
`Topical Review
`
`1. Background to the topical review
`
`Photoplethysmography (PPG) is an optical measurement technique that can be used to detect
`blood volume changes in the microvascular bed of tissue (Challoner 1979). It has widespread
`clinical application, with the technology utilized in commercially available medical devices,
`for example in pulse oximeters, vascular diagnostics and digital beat-to-beat blood pressure
`measurement systems. The basic form of PPG technology requires only a few opto-electronic
`components: a light source to illuminate the tissue (e.g. skin), and a photodetector to measure
`the small variations in light intensity associated with changes in perfusion in the catchment
`volume. PPG is most often employed non-invasively and operates at a red or a near
`infrared wavelength. The most recognized waveform feature is the peripheral pulse, and it is
`synchronized to each heartbeat. Despite its simplicity the origins of the different components
`of the PPG signal are still not fully understood. It is generally accepted, however, that they
`can provide valuable information about the cardiovascular system (Kamal et al 1989).
`This review has two parts. An introductory section describes the basic principle of PPG
`operation, light interaction with tissue, early and recent history of PPG, instrumentation,
`measurement protocol, and pulse wave analysis. The second section reviews current and
`potential clinical applications in physiological measurement under the categories of clinical
`physiological monitoring, vascular assessment and autonomic function.
`
`2. Photoplethysmography
`
`2.1. The photoplethysmography waveform
`
`The pulsatile component of the PPG waveform is often called the ‘AC’ component and usually
`has its fundamental frequency, typically around 1 Hz, depending on heart rate (figure 1).
`This AC component is superimposed onto a large quasi-DC component that relates to the
`tissues and to the average blood volume. This DC component varies slowly due to respiration,
`vasomotor activity and vasoconstrictor waves, Traube Hering Mayer (THM) waves and
`also thermoregulation (Burton 1939, Burton and Taylor 1940, Hertzman and Dillon 1940b,
`Hertzman and Roth 1942a, 1942b, 1942c, Hertzman and Flath 1963, Hyndman et al 1971,
`Pe˜n´az 1978, Ahmed et al 1982, Harness and Marjanovic 1989, Nitzan et al 1996b, 1996a,
`general thermoregulatory pulse changes in Shusterman et al (1997), Schultz-Ehrenburg and
`Blazek (2001), Nitzan et al (2001)). These characteristics are also body site dependent (Allen
`and Murray 2000b). With suitable electronic filtering and amplification both the AC and DC
`can be extracted for subsequent pulse wave analysis.
`
`2.2. Optical considerations of the origins of the photoplethysmography waveform
`
`The interaction of light with biological tissue is complex and includes the optical processes
`of (multiple) scattering, absorption, reflection, transmission and fluorescence (Anderson and
`Parrish 1981). Several researchers have investigated the optical processes in relation to PPG
`measurements (Hertzman and Randall 1948, Brown et al 1965, D’Agrosa and Hertzman 1967,
`Weinman 1967, Zweifler et al 1967, Challoner 1979, Ochoa and Ohara 1980, Nijboer et al
`1981, Roberts 1982, Lindberg and ¨Oberg 1993, de Trafford and Lafferty 1984, Kamal et al
`1989). They have highlighted the key factors that can affect the amount of light received
`by the photodetector: the blood volume, blood vessel wall movement and the orientation of
`red blood cells (RBC). The orientation effect has been demonstrated by recording pulsatile
`waveforms from dental pulp and in a glass tube where volumetric changes should not be
`
`FITBIT, Ex. 1021
`
`

`

`Topical Review
`
`R3
`
`Figure 1. The pulsatile (AC) component of the PPG signal and corresponding electrocardiogram
`(ECG). The AC component is actually superimposed on a much larger quasi-DC component that
`relates to the tissues and to the average blood volume within the sample. It represents the increased
`light attenuation associated with the increase in microvascular blood volume with each heartbeat.
`In practice, the PPG waveform is often inverted.
`
`possible, and more recently by N¨aslund et al (2006) who detected pulsatile waveforms in
`bone. The recorded pulses do bear a direct relationship with perfusion, and the greater the
`blood volume the more the light source is attenuated. However, attempts at pulse amplitude
`quantification (‘calibration’) have been largely unsuccessful (Hertzman 1938, Challoner and
`Ramsay 1974, Jespersen and Pedersen 1986, Cejnar et al 1993).
`The wavelength of optical radiation is also important in light–tissue interactions (Cui et al
`1990), and for three main reasons: (1) The optical water window: the main constituent of
`tissue is water that absorbs light very strongly in the ultraviolet and the longer infrared
`wavelengths. The shorter wavelengths of light are also strongly absorbed by melanin. There
`is, however, a window in the absorption spectra of water that allows visible (red) and near
`infrared light to pass more easily, thereby facilitating the measurement of blood flow or
`volume at these wavelengths. Thus, the red or near infrared wavelengths are often chosen
`for the PPG light source (Jones 1987), (2) Isobestic wavelength: significant differences exist
`in absorption between oxyhaemoglobin (HbO2) and reduced haemoglobin (Hb) except at
`the isobestic wavelengths (Gordy and Drabkin 1957). For measurements performed at an
`isobestic wavelength (i.e. close to 805 nm, for near infrared range) the signal should be largely
`unaffected by changes in blood oxygen saturation, and (3) Tissue penetration depth: the depth
`to which light penetrates the tissue for a given intensity of optical radiation depends on the
`operating wavelength (Murray and Marjanovic 1997). In PPG the catchment (study) volume,
`depending on the probe design, can be of the order of 1 cm3 for transmission mode systems.
`PPG can provide information about capillary nutritional blood flow and the thermoregulatory
`blood flow through arterio-venous anastomosis shunt vessels.
`
`2.3. Early and recent history of photoplethysmography
`
`This paragraph gives a brief summary of the early history of PPG and is taken from the
`excellent review article by Challoner (1979).
`In 1936 two research groups (Molitor and
`Kniazuk of the Merck Institute of Therapeutic Research, New Jersey, and Hanzlik et al of
`Stanford University School of Medicine) described similar instruments used to monitor the
`blood volume changes in the rabbit ear following venous occlusion and with administration
`of vasoactive drugs. Molitor and Kniazuk also described recordings made with a reflection
`mode PPG system from human fingers. A pioneer who helped establish the PPG technique
`was Alrick Hertzman from the Department of Physiology at St. Louis University School of
`
`FITBIT, Ex. 1021
`
`

`

`R4
`
`Topical Review
`
`Medicine, St. Louis, MO. In 1937, Hertzman and his colleagues published their first paper
`on PPG describing the use of a reflection mode system to measure blood volume changes
`in the fingers induced by the Valsalva manoeuvre, exercise and with exposure to cold. This
`excellent contribution to the field demonstrated the potential clinical utility of the technique.
`In 1938, Hertzman undertook a validation of the PPG technique by comparing blood volume
`changes with those measured simultaneously by mechanical plethysmography. Preliminary
`observations on the PPG technique were also reported in the same year by Matthes and
`Hauss. Hertzman and Dillon (1940a) split the AC and DC components with separate electronic
`amplifiers and monitored vasomotor activity. Potential sources of error with the technique have
`been identified by Hertzman (1938), who emphasized that good contact with skin was needed,
`but without excessive pressure that would result in blanching. He advised that movement
`of the measurement probe against the skin should be avoided. These observations led to the
`development of elaborate positioning devices. Illumination was identified as another important
`design consideration. Hertzman also used a battery powered torch bulb which was less than
`ideal because of its relatively wide spectrum, particularly in the infrared because of local tissue
`heating, errors due to the effects of oxygen saturation, and the widespread illumination which
`can mix skin microvascular blood flow with larger vessel signals. Furthermore, constant light
`intensity could not be guaranteed.
`In more recent decades the desire for small, reliable, low-cost and simple-to-use non-
`invasive (cardiovascular) assessment techniques are key factors that have helped re-establish
`photoplethysmography. Advances in opto-electronics and clinical instrumentation have also
`significantly contributed to its advancement. The developments in semiconductor technology,
`i.e. light emitting diodes (LED), photodiodes and phototransistors, have made considerable
`improvements in the size, sensitivity, reliability and reproducibility of PPG probe design. A
`major advance in the clinical use of a PPG-based technology came with the introduction of the
`pulse oximeter as a non-invasive method for monitoring patients’ arterial oxygen saturation
`(Aoyagi et al 1974, Yoshiya et al 1980). There have also been considerable developments in
`computer-based digital signal processing and pulse wave analysis.
`
`2.4. Photoplethysmography instrumentation
`
`Modern PPG sensors often utilize low cost semiconductor technology with LED and matched
`photodetector devices working at the red and/or near infrared wavelengths (CIE IR-A near
`infrared band 0.8 to 1 μm, Duck (1990)). An excellent review of optical sensor technology
`for PPG and pulse oximetry applications is written by Webster (1997).
`(cid:2)(cid:2)
`berg
`The choice of light source is important (Burke and Whelan 1986, Lindberg and O
`1991, Ugnell and ¨Oberg 1995). LEDs convert electrical energy into light energy and have a
`narrow single bandwidth (typically 50 nm). They are compact, have a very long operating
`life (>105 h), operate over a wide temperature range with small shifts in the peak-emitted
`wavelength, and are mechanically robust and reliable. The averaged intensity of the LED
`should be constant and preferably be sufficiently low to minimize excessive local tissue heating
`and also reduce the risk of a non-ionizing radiation hazard. The choice of photodetector is
`also important (Weinman and Fine 1972, Fine and Weinman 1973). Its spectral characteristics
`are chosen to match that of the light source. A photodetector converts light energy into an
`electrical current. They are compact, low-cost, sensitive, and have fast response times. Near
`infrared devices can be encased with daylight filters. The photodetector connects to low noise
`electronic circuitry that includes a transimpedance amplifier and filtering circuitry.
`A high pass filter reduces the size of the dominant DC component and enables the
`pulsatile AC component to be boosted to a nominal 1 V peak-to-peak level. Carefully
`
`FITBIT, Ex. 1021
`
`

`

`Topical Review
`
`R5
`
`(a)
`
`(b)
`
`Figure 2. Electronic building blocks used in a typical PPG measurement system.
`(a) A
`transimpedance (current-to-voltage) amplifier stage that converts light intensity at the photodiode
`(PD) to an amplifier output voltage (V = I × R, transimpedance gain proportional to feedback
`resistor value R). (b) The signal conditioning stages surrounding the transimpedance amplifier
`which include low pass filtering, high pass filtering and further amplification, inversion and signal
`interfaces. The AC component and a measure of the DC component are available for pulse wave
`analysis. A constant current driver stage for the PPG LED is also shown.
`
`chosen filtering circuitry is also needed to remove the unwanted higher frequency noise such
`as electrical pick up from (50 Hz) mains electricity frequency interference. Figure 2(a)
`shows a transimpedance amplifier design and figure 2(b) shows the signal conditioning stages
`surrounding this, including low pass filtering, high pass filtering and further amplification,
`signal inversion and signal interface. The choice of high pass filter cut-off frequency is
`particularly important and is often a design compromise; excessive filtering can distort the
`pulse shape but too little filtering can result in the quasi-DC component dominating over the
`AC pulse (Allen and Murray 2003, 2004). This example system shows a constant current
`driver stage for the PPG probe LED.
`There are two main PPG operational configurations: transmission (‘trans-illumination’)
`mode operation where the tissue sample (e.g. fingertip) is placed between the source and
`detector, and reflection (‘adjacent’) mode operation where the LED and detector are placed
`side-by-side. Clearly, transmission mode PPG imposes more restrictions than the reflection
`mode PPG on the body locations available for study. The PPG probe should be held securely
`in place to minimize probe-tissue movement artefact. There are other sources of artefact that
`need to be considered in the measurement technology. For example, artefact can arise from
`ambient light interference but can be reduced in several ways: by suitable probe attachment
`to the skin (e.g. using a dark Velcro wrap-around cuff), by further shading of the study site
`area and performing measurements in subdued lighting, and by electronic filtering (e.g. light
`modulation filtering, Webster (1997)). Ambient light interference in PPG-based systems has
`also been discussed by Hanowell et al (1987).
`Many of the studies reported in the PPG literature are for a single site, often the ear, finger
`or toe, where pulses can easily be detected (including Stern (1974), Barnes et al (1977a, 1977b),
`Sherebrin and Sherebrin (1990), Allen and Murray (1993), Chowienczyk et al (1999), Hahn
`et al (1999), Bortolotto et al (2000), Foo et al (2006), Millasseau et al (2006)). Many other
`skin measurement sites are available for vascular assessment (Tur et al 1983). The supraorbital
`
`FITBIT, Ex. 1021
`
`

`

`R6
`
`Topical Review
`
`artery, just above each eye has been studied (Lee et al 1981), as well as measurements on the
`skin above key arterial landmarks (Loukogeorgakis et al 2002). There have also been studies
`published which investigated PPG pulses at two sites simultaneously (Dillon and Hertzman
`1941, Nijboer and Dorlas 1985, Cooke et al 1985, Grossmann et al 1987, Okada et al 1986,
`Porret et al 1995, Nitzan et al 1998b, 1998a, 2001). Three channel PPG research systems
`have been described by Evans and Geddes (1988) and four channel research systems by Buchs
`et al (2005) and Erts et al (2005). Multiple finger site pulse data have also been reported by
`Dyszkiewicz and Tendera (2006). Six channel PPG data have been published for simultaneous
`multi-bilateral site PPG data measurements (i.e. the right and left ear lobes, index fingers or
`thumbs, and great toes) (Allen et al 2000, 2002, 2003, 2004, 2005, 2006). Technically, Allen
`and Murray (2000a) have emphasized the importance of the electronic and optical matching of
`pulse amplifier channels to allow the best chance for right-to-left side physiological differences
`to be detected with confidence. A multi-bilateral site PPG pulse measurement and analysis
`system and example pulse recording from a healthy subject are shown in figure 3. PPG systems
`are also available commercially. Examples include the Skidmore Medical Ltd. Vicorder, the
`Cuban Biof´ısica M´edica ANGIODIN R(cid:3)
`PD 3000, and the VIASYS Healthcare MicroLiteTM
`and VasoGuardTM systems. These lists are not exhaustive.
`Other emerging technologies include PPG imaging technology, telemedicine and remote
`monitoring.
`Schultz-Ehrenburg and Blazek (2001) and Huelsbusch and Blazek (2002)
`investigated an experimental cooled near infrared CCD PPG imaging system for studying
`skin blood flow and related rhythmical phenomena. The aim of the technology was to obtain
`new insights into normal physiological tissue perfusion and detect changes associated with
`ulcer formation and wound healing. In 2005, Wieringa et al described a contactless multiple
`wavelength PPG imaging system whose main application is the remote imaging of arterial
`oxygen saturation (SpO2) distribution within tissue. The system acquires movies of two-
`dimensional matrix spatially resolved PPG signals at three wavelengths (660 nm, 810 nm
`and 940 nm) during changes in respiration. A tissue oxygen image might be useful in
`many areas of medical diagnostics, for example in quantifying tissue viability. PPG has
`considerable potential for telemedicine including the remote/home health monitoring of
`patients. Miniaturization, ease-of-use and robustness are key design requirements for such
`systems. This is illustrated with finger ring-based PPG sensors for monitoring beat-to-beat
`pulsations (Rhee et al 2001, Zheng and Zhang 2003) and the need for motion artefact reduction,
`optimal sensor placement and minimizing battery power consumption. Innovative LED and
`photodetector array technology has been incorporated into a PPG finger sensor and palm-sized
`home health monitor to enable the pulse, oxygen saturation and respiration to be measured
`along with haematocrit derived from optical characteristics at five different wavelengths (569,
`660, 805, 904 and 975 nm) (Yoon et al 2005). Preliminary clinical testing showed that
`the haematocrit was within ±10% of the gold standard value. Respiratory information was
`extracted using digital filtering techniques and blood oxygen saturation (SpO2) predicted using
`the standard ratio for the red and near infrared wavelengths.
`
`2.5. Measurement protocol and reproducibility
`Reproducibility is very important in clinical physiological measurement, for example in giving
`confidence in detecting significant responses to therapy. Many factors affect reproducibility,
`including the method of probe attachment to tissue, probe–tissue interface pressure, pulse
`amplifier bandwidth, minimization of movement artefact, subject posture and relaxation,
`breathing, wakefulness, room temperature and acclimatization (Teng and Zhang 2006, Zhang
`and Zhang 2006). As yet, however, there are no internationally recognized standards for
`clinical PPG measurement. Published research tends to be based on studies using quite
`
`FITBIT, Ex. 1021
`
`

`

`Topical Review
`
`(a)
`
`R7
`
`(b)
`
`Figure 3. Multi-bilateral site photoplethysmography.
`(a) An overview of a six channel PPG
`measurement and analysis system, giving capability for pulses to be compared between the right
`and left sides and between head to foot sites (Allen et al 2006). (b) An example recording made
`from the right and left ear lobes, index fingers, and great toes of a healthy subject. There is similarity
`in the pulse characteristics between the right and left body sides but clear differences between the
`proximal and distal measurement sites. However, the degree of right to left side similarity in the
`high and low frequency components of the PPG waveform can be reduced in patients with vascular
`disease (Allen and Murray 2000a) and also in patients with autonomic dysfunction (Buchs et al
`2005).
`
`FITBIT, Ex. 1021
`
`

`

`R8
`
`Topical Review
`
`differing measurement technology and protocols, thereby limiting the ease with which PPG
`physiological measurements can be replicated between research centres.
`There are a limited number of studies quantifying the repeatability or reproducibility of
`PPG measurements. An important study by Jago and Murray (1988) addressed the uncertainty
`in PPG measurements for a group of healthy adult subjects. They studied the repeatability of
`PPG pulse transit time (PTT) measurements made from the ear, thumb and toe sites both within
`session and between sessions held on separate days. Measurements at individual sites and
`bilateral (right–left) side differences were both assessed. The results showed the importance of
`controlling for factors such as posture, ambient temperature, relaxation and acclimatization.
`Bilateral measurements were generally more repeatable than individual site measurements
`since heart rate, respiration and blood pressure factors tend to affect both sides of the body
`simultaneously.
`There have also been a limited number of studies published that quantify the complex
`physiological variability in PPG waveforms measured at different body sites. Applications that
`utilize the beat-to-beat variation in PPG characteristics include the assessment of autonomic
`dysfunction and cardiovascular ageing (see section 3.3). It can also be useful, however, to
`obtain an averaged pulse measure to represent an individual subject/site. An averaging period
`covering at least 60 heartbeats has been suggested to improve confidence in the single timing,
`amplitude or shape measurements extracted from the PPG pulse (Allen 2002).
`
`2.6. Photoplethysmography pulse wave characterization and analysis
`
`2.6.1. Pulse wave characterization. Two important characteristics of the PPG AC pulse
`waveform were described by Hertzman and Spealman (1937). The appearance of the pulse
`was defined as two phases: the anacrotic phase being the rising edge of the pulse, and the
`catacrotic phase being the falling edge of the pulse. The first phase is primarily concerned
`with systole, and the second phase with diastole and wave reflections from the periphery. A
`dicrotic notch is usually seen in catacrotic phase of subjects with healthy compliant arteries.
`It is useful also to consider the blood pressure pulse and its propagation along individual
`arteries. The pressure pulse wave is known to change in shape as it moves toward the periphery
`and undergoes amplification and alterations in its shape and temporal characteristics. These
`changes are thought to be largely due to reflection of the pulse wave and the tapering down
`of the arteries towards the periphery. Pulse propagation in arteries is further complicated by
`frequency dependent phase distortion. These phenomena have been described by O’Rourke
`and Gallagher (1996) and are discussed in the wider literature on pulse. The blood pressure
`pulse has similarities with the PPG blood volume pulse, with similar changes occurring in
`vascular disease, such as damping and a loss of pulsatility. The damping has been associated
`with a reduction in vessel compliance and increased peripheral resistance, although these
`changes have yet to be fully explained.
`The potential of PPG for assessing vascular disease was recognized many decades ago. In
`1940b, Hertzman and Dillon compared PPG to mechanical plethysmography in arteriopaths
`and in normal control subjects. They derived a crest time measurement from the rising edge
`of the pulse waveform and normalized this to the heart rate. The crest time was prolonged
`in patients with vascular disease or hypertension. The potential for extracting diagnostic
`information from the PPG pulse has been reviewed by Murray and Foster (1996). From the
`literature, many features have been investigated (figure 4), including beat-to-beat PPG rise
`time, PTT, amplitude, shape, and the variability in each of these. The pulse shape (contour)
`can also be described after normalization in pulse width and height (Allen and Murray 2003).
`
`FITBIT, Ex. 1021
`
`

`

`Topical Review
`
`(a)
`
`R9
`
`(b)
`
`1.2
`
`1
`
`0.8
`
`0.6
`
`0.4
`
`0.2
`
`0
`
`0
`
`1.2
`
`1
`
`0.8
`
`0.6
`
`0.4
`
`0.2
`
`0
`
`0
`
`1
`
`0.8
`
`0.6
`
`0.4
`
`0.2
`
`0
`
`0
`
`20
`
`40
`
`60
`
`80
`
`100
`
`20
`
`40
`
`60
`
`80
`
`100
`
`20
`
`40
`
`60
`
`80
`
`100
`
`1.2
`
`1
`
`0.8
`
`0.6
`
`0.4
`
`0.2
`
`0
`
`0
`
`1.2
`
`1
`
`0.8
`
`0.6
`
`0.4
`
`0.2
`
`0
`
`0
`
`1
`
`0.8
`
`0.6
`
`0.4
`
`0.2
`
`0
`
`0
`
`20
`
`40
`
`60
`
`80
`
`100
`
`20
`
`40
`
`60
`
`80
`
`100
`
`20
`
`40
`
`60
`
`80
`
`100
`
`Figure 4. Characterization of PPG pulse timing, amplitude and shape features. (a) Key pulse
`landmarks can automatically be identified using a pulse wave analysis computer to give beat-to-
`beat pulse transit time to the foot of the pulse (PTTf), pulse transit time to the peak of the pulse
`(PTTp), and foot-to-peak amplitude (AMP). The pulse landmarks can then be used to calculate
`the normalized pulse contour. Contour examples are given in (b) for two different healthy subjects
`(Allen and Murray 2003).
`
`FITBIT, Ex. 1021
`
`

`

`R10
`
`Topical Review
`
`(a)
`
`(b)
`
`(c)
`
`(d)
`
`Figure 5. Examples of the types of measurement artefact and the extremes in physiological
`variation that can be seen in PPG recordings. Each recording is from the index finger site over a
`period of 1 min and the artefact/physiological events are marked. (a) An episode of gross movement
`artefact or PPG probe cable tugging lasting approximately 15 s. (b) Hand or finger tremor, (c) a
`bout of coughing, and (d) marked changes in the breathing pattern (a deep gasp or yawn). These
`types of artefact and physiological variation should be considered with the measurement protocol
`and subsequent pulse wave analysis.
`
`2.6.2. Pulse wave analysis. Manual measurement and feature extraction techniques were
`often used in the very early days of pulse wave analysis, using various media including
`chart recorder paper and ruler or photographic recording/magnetic tape (e.g. Hertzman and
`Spealman (1937), Dillon and Hertzman (1941), Simonson (1956), Corte (1979), Cooke et al
`(1985), Sherebrin and Sherebrin (1990)). Recent developments in computing technology
`and software data analysis tools have enabled the sophisticated pre- and post-processing
`of physiological waveforms. MATLAB (MathWorks Inc.) is a digital signal processing
`environment that is well suited to pulse wave analysis algorithm prototyping, and often
`appears in the PPG literature.
`It is well established that PPG measurements are quite sensitive to patient and/or probe–
`tissue movement artefact (see examples in figure 5). The automatic detection of such motion
`
`FITBIT, Ex. 1021
`
`

`

`Topical Review
`
`R11
`
`artefact, and its separation from good quality although highly variable pulse recordings,
`is a non-trivial exercise in computer signal processing. Computer-based filtering, feature
`extraction and waveform averaging have also been employed in PPG pulse wave analysis,
`including the analysis of frequency (de Trafford et al 1982, Okada et al 1986, Nitzan et al
`1994, Bernardi et al 1996, Grohmann et al 1996a, 1996b, Larsen et al 1997, Sherebrin
`and Sherebrin 1990), joint-time frequency (Yan et al 2005), artificial neural network (Allen
`and Murray 1993, 1995, 1996, 1999, Weng et al 1998), systems identification and transfer
`function modelling (Cohn et al 1995, Allen and Murray 1993, 1995, 1996, McVeigh et al
`1999, Millasseau et al 2000), principal component analysis (Enr´ıquez et al 2002), nonlinear
`and chaos theory (Christ et al 1997, Bhattacharya et al 2001), cross correlation (Allen and
`Murray 2000a, Drinnan et al 2001) and double differentiation (acceleration plethysmogram,
`Takada et al (1996–97), Takazawa et al (1998), Bortlotto et al (2000)).
`
`3. Clinical applications
`
`PPG has been applied in many different clinical settings, including clinical physiological
`monitoring (blood oxygen saturation, heart rate, blood pressure, cardiac output and
`respiration), vascular assessment
`(arterial disease,
`arterial compliance and ageing,
`endothelial function, venous assessment, vasospastic conditions, e.g. Raynaud’s phenomenon,
`microvascular blood flow and tissue viability) and autonomic function (vasomotor function and
`thermoregulation, blood pressure and heart rate variability, orthostatic intolerance, neurology
`and other cardiovascular variability assessments). This section reviews each of these areas
`with a view to demonstrating the widespread use of the optical technology in medicine and
`also its considerable potential for further innovation and application.
`
`3.1. Clinical physiological monitoring
`
`3.1.1. Blood oxygen saturation. Pulse oximetry is said to represent one of the most significant
`technological advances in clinical patient monitoring over the last few decades (Webster
`1997). It utilizes PPG measurements to obtain information about the arterial blood oxygen
`It has widespread
`saturation (SpO2) as well as heart rate (Aoyagi and Miyasaka 2002).
`application in many different clinical settings, including hospital, outpatient, sports medicine,
`domiciliary use, and in veterinary clinics.
`In the early 1990s pulse oximetry became a
`mandated international standard for monitoring during anaesthesia. An excellent review of
`the technique can be found in Kyriacou (2006) where he described the basic principle of
`operation, measurement technology and its clinical applications. Earlier reviews of pulse
`oximetry have been written by Kelleher (1989) and Severinghaus and Kelleher (1992).
`SpO2 can be determined by shining red and then near infrared light through vascular tissue,
`with rapid switching between the wavelengths. The amplitudes of the red and near infrared
`AC signals are sensitive to changes in SpO2 because of the differences in the light absorption
`of HbO2 and Hb at these two wavelengths. From their amplitude ratio, and corresponding
`PPG DC components, SpO2 can be estimated. This process usually involves an empirically
`derived calibration factor (Webster 1997). There is the assumption that the pulsatile component
`of the PPG signal results solely from the arterial blood volume changes with each heartbeat.
`Limitations of pulse oximetry are that the technique relies upon a peripheral pulse to be present,
`oxygen saturation readings can be affected by dyshaemoglobinaemias, and its accuracy can
`fall off at low saturation levels (Kyriacou 2006). Advanced computer algorithms have been
`developed to try to overcome the problem of movement artefact affecting the measurement
`
`FITBIT, Ex. 1021
`
`

`

`R12
`
`Topical Review
`
`reliability. This includes the Masimo SET technology (Goldman et al 2000, Hayes and Smith
`2001).
`Pulse oximeters can measure SpO2 using both the reflection and transmission modes
`of operation (Mendelson and Ochs 1988). Central body measurement sites have also been
`investigated, including oesophageal oxygen saturation monitoring to overcome the problem of
`finger pulse loss associated with intra-operative peripheral cooling (Kyriacou et al 2002,
`Kyriacou 2006). Another recent and exciting development in pulse oximetry is the non-invasive
`measurement of venous oxygen saturation using external artificial perturbations applied close
`to the PPG probe (VENOX Technology, Chan et al (200