HK1227666B - System for measurement of cardiac output, stroke volume, cardiac power, and blood pressure - Google Patents

Description

Measurement systems for cardiac output, stroke volume, cardiac effort, and blood pressure

This application is a divisional application of the Chinese national phase patent application with application number 201180066468.5, which entered the Chinese national phase on August 1, 2013 and has the international application number PCT/US2011/067441 and the international application date of December 27, 2011, and the invention name is "Body-worn system for continuous and non-invasive measurement of cardiac output, stroke volume, cardiac force and blood pressure".

CROSS-REFERENCE TO RELATED APPLICATIONS

This application claims priority to U.S. provisional application serial number 61/427,756, filed December 28, 2010, entitled “BODY-WORN SYSTEM FOR CONTINUOUS, NONINVASIVE MEASUREMENT OF CARDIAC OUTPUT, STROKEVOLUME, AND BLOOD PRESSURE,” which is hereby incorporated by reference in its entirety.

STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT

Work on some components described in this patent application was supported by the U.S. Department of Defense under Contract W81XWH-11-2-0085.

Background of the Invention

Field of the Invention

The present invention relates to medical devices for monitoring cardiovascular performance, such as cardiac output (CO), stroke volume (SV), and continuous non-invasive blood pressure (cNIBP).

Related technical notes

CO is typically measured in a hospital setting and informally indicates how effectively a patient's heart is pumping blood through their arterial tree. More specifically, CO (measured in liters per minute) describes the time-dependent volume of blood ejected from the left ventricle into the aorta; it indicates how well a patient's heart is delivering blood-borne oxygen, nutrients, and other substances to the body's cells. CO is the product of heart rate (HR) and SV, where SV is defined as the mathematical difference between the left ventricular end-diastolic volume (EDV) and the end-systolic volume (ESV), i.e.:

CO＝SV×HR (1)

Combining CO with mean arterial blood pressure (MAP) into a single value, termed "cardiac power" (CP), provides a particularly valuable prognostic variable for monitoring patients with cardiac conditions such as congestive heart failure (CHF) and is an independent predictor of mortality that can be measured noninvasively using cardiopulmonary exercise testing. Specifically, CP is defined as:

CP＝CO×MAP (2)

Measuring CO and SV continuously and noninvasively with high clinical accuracy has generally been considered the holy grail of medical device monitoring. Most existing technologies in this field require an indwelling catheter (which could potentially harm the patient), are inherently inaccurate in critical situations, and require specially trained operators. For example, the current "gold standards" for this measurement are thermodilution cardiac output (TDCO) and the Fick oxygen principle (Fick). However, both TDCO and Fick are highly invasive techniques that can cause infection and other complications, even in a carefully controlled hospital environment. With TDCO, a pulmonary artery catheter (PAC) (also known as a Swan-Ganz catheter) is typically inserted into the right portion of the patient's heart. Procedurally, a bolus (typically 10 ml) of glucose or saline (chilled to a known temperature) is infused through the PAC. A temperature measurement device within the PAC (positioned at a known distance (typically 6 to 10 cm) from the point of infusion) measures the gradually increasing temperature of the diluted blood. CO is then estimated from the measured time-temperature curve (known as the "thermodilution curve"). The larger the area under this curve, the lower the cardiac output. Similarly, a smaller area under the curve means a shorter conduction time for the cold bolus to dissipate, and thus a higher CO.

Fick involves calculating the amount of oxygen consumed and distributed throughout a patient's blood over a given period of time. The algorithm associated with the technique incorporates oxygen consumption (as measured by a spirometer) with the difference between the oxygen content of central blood measured from a PAC and the oxygen content of peripheral arterial blood measured from an indwelling cannula.

Both TD and Fick typically measure CO to between about 0.5 l/min and 1.0 l/min, or with an accuracy of about +/- 20%, during critical illness.

Several noninvasive techniques for measuring SV/CO/CP have been developed in the hope of overcoming the shortcomings of Fick and TD. For example, Doppler-based ultrasound echoes (Doppler/ultrasound) use the well-known Doppler frequency shift to measure blood flow velocity and have been shown to be reasonably accurate compared to more invasive methods. Both two-dimensional and three-dimensional versions of this technique require specially trained operators and are therefore impractical for continuous measurement, with the exception of transesophageal Doppler techniques. CO/SV can also be measured by techniques that rely on electrodes placed on the patient's torso, which inject and then collect a low-ampere, high-frequency modulated current. These techniques (based on bioelectrical impedance and known as "impedance cardiography" (ICG), "electrical cardiovelocimetry" (ECV), and "bioreactance" (BR)) measure time-dependent electrical waveforms modulated by blood flowing through the patient's chest. Blood is a good electrical conductor and, when pumped by the heart, can further modulate the current injected by these techniques in a manner sensitive to the patient's CO. During measurement, ICG, ECV, and BR each extract properties called left ventricular ejection time (LVET) and pre-ejection period (PEP) from the time-dependent ICG and ECG waveforms. A processor then analyzes the waveforms using an empirical mathematical equation (shown below in Equation 2) to estimate SV. CO is determined by multiplying SV by HR, as described above in Equation 1.

ICG, ECV and BR all represent continuous, non-invasive alternatives for measuring CO/SV and can, in theory, be performed with inexpensive systems and without specially trained operators. However, the medical community has not yet accepted the methods, despite the fact that clinical studies have shown that they are effective for certain patient populations. In 1992, for example, an analysis by Fuller et al. analyzed data from 75 published studies describing the correlation between ICG and TD/Fick (Fuller et al., The validity of cardiac output measurement by thoracic impedance: a meta-analysis; Clinical Investigative Medicine; 15: 103-112 (1992)). The study concluded using a meta-analysis that, in 28 of these studies, ICG showed a correlation between dye dilution and FickCO (r = 0.80-0.83 for TDCO). Patients classified as critically ill (e.g., those with acute myocardial infarction, sepsis, and pulmonary fluid overload) had worse outcomes. Further hindering the commercial acceptance of these technologies is the relatively large size and complexity of ICG monitors compared to conventional vital signs monitors, which tend to be similar. This means that two large and expensive pieces of monitoring equipment may need to be positioned at the bedside to monitor a patient's vital signs and CO/SV. For this and other reasons, impedance-based CO measurement has not yet achieved widespread commercial success.

ICG-based methods for measuring CO/SV have evolved since Fuller's analysis. For example, recent studies have shown that the dimensionless peak rate of change of the transthoracic electrical impedance pulse, defined as the _maximum value of the derivative of the ICG waveform (dZ/dt) divided by the base impedance ( _Zo ), is an analog of acceleration (in units of 1/ ^s2 ). When subjected to a square root transformation, this yields the ohmic mean velocity [(dZ/dt) _max / _Zo )] ^0.5 . These parameters are described in detail in U.S. Patents 7,740,590 and 7,261,697, the entire contents of which are incorporated herein by reference. When this value is multiplied by LVET and volume conductivity ( _Vc ), which is related to body shape variation with intrathoracic blood volume through body weight, a reasonable approximation of SV is obtained. Compared with CO measured by TDCO and transesophageal echocardiography, good to high correlation and a limit of agreement within +/-30% have been reported.

Although most ICG measurements are made on the thorax, there is good evidence in the literature suggesting that left ventricular SV can be obtained from the upper extremities, and particularly the brachial artery. For example, Chemla et al. showed that peak aortic blood flow acceleration correlated well with peak brachial artery blood flow acceleration (r = 0.79) (see, e.g., Chemla et al., Blood flow acceleration in the carotid and brachial arteries of healthy volunteers: respective contributions of cardiac performance and local resistance; Fundam Clin Pharmacol; 10:393-399 (1996)). This study also demonstrated that, while brachial blood flow velocity is influenced by downstream vasoactivity, peak brachial blood flow acceleration is influenced only by upstream beta-adrenergic influences from cardiac beat formation. This suggests that the square root transformation of the brachial (dZ/dt) _max / _Zo can yield an accurate estimate of SV when multiplied by LVET and Vc of appropriate magnitude. Stanley et al. showed that the slope of the maximum early systolic rise of transthoracic and brachial impedance changes (ΔZ) was identical, indicating that they were linearly related (see, e.g., Stanley et al., Multi-channel electrical bioimpedance: a new noninvasive method to simultaneously measure cardiac and peripheral blood flow; J Clin Monit Comput; 21:345-51 (2007)). This means that, despite having different amplitudes, the peak rates of change of transthoracic and transbrachial impedance changes can both be used to calculate SV. Finally, Wang et al. demonstrated that impedance changes in the forearm (ΔZ(t)) were highly correlated with Doppler-derived SV, showing a correlation coefficient of r = 0.86 (see, e.g., Wang et al., Evaluation of changes in cardiac output from the electrical impedance waveform in the forearm; Physiol Meas; 28:989-999 (2007)).

CO/SV can also be estimated from time-dependent arterial blood pressure waveforms, measured, for example, by a sphygmomanometer or an indwelling arterial catheter. Algorithms can be used to extract pulse pressure (PP) and other profile-related features from these waveforms, which are then processed to estimate CO/SV. Unfortunately, the heart and its associated vasculature can function independently and sometimes in conflict, so changes in parameters like PP can both reflect and mask changes in CO/SV. In other words, CO measurements using time-dependent arterial waveforms represent a combination of cardiac and vascular function.

Pulse arrival time (PAT), defined as the transit time of a pressure pulse initiated by a heartbeat through a patient's arterial system, has been shown in multiple studies to correlate with both systolic (SYS) and diastolic (DIA) blood pressure. In these studies, PAT is typically measured using a conventional vital signs monitor that includes separate modules for determining electrocardiogram (ECG) and pulse oximetry (SpO2) values. During a PAT measurement, multiple electrodes are typically attached to the patient's chest to measure the time-dependent components of the ECG waveform, which are characterized by spikes known as "QRS complexes." The QRS complex indicates the initial depolarization of the ventricles within the heart and informally marks the onset of the heartbeat and the subsequent pressure pulse. SpO2 is typically measured using a bandage or clothespin-shaped sensor that attaches to the patient's finger and includes an optical system operating in both the red and infrared spectral regions. A photodetector measures radiation emitted from the optical system that is transmitted through the patient's finger. Other body parts (e.g., ear, forehead, and nose) can also be used instead of a finger. During measurement, a microprocessor analyzes both the red and infrared radiation measured by the photodetector to determine time-dependent waveforms corresponding to different wavelengths, each referred to as a photoplethysmogram (PPG) waveform. SpO2 values are calculated from these waveforms. The time-dependent characteristics of the PPG waveform indicate both pulse rate and volume absorption rate changes in the underlying artery (e.g., in the finger) caused by the propagating pressure pulse.

A typical PAT measurement measures the time between the maximum point on the QRS complex (indicating the peak of ventricular depolarization) and a portion of the PPG waveform (indicating the arrival of the pressure pulse). PAT depends primarily on arterial compliance, the propagation distance of the pressure pulse (which is closely approximated by the patient's arm length), and blood pressure. To account for patient-specific properties such as arterial compliance, PAT-based measurements of blood pressure are typically "calibrated" using a conventional blood pressure cuff. Typically, during the calibration process, a blood pressure cuff is applied to the patient, used to take one or more blood pressure measurements, and then removed. Next, calibration measurements are used, along with changes in PAT, to determine the patient's blood pressure and blood pressure variability. PAT is generally inversely correlated with blood pressure, i.e., a decrease in PAT indicates an increase in blood pressure.

Many issued U.S. patents describe the relationship between PAT and blood pressure. For example, U.S. Patents 5,316,008; 5,857,975; 5,865,755; and 5,649,543 each describe a device comprising conventional sensors that measure ECG and PPG waveforms, which are then processed to determine PAT.

SUMMARY OF THE INVENTION

The present invention provides a small, body-worn monitor for measuring SV/CO/CP, along with cNIBP, HR, respiratory rate (RR), SpO2, as well as body temperature (TEMP), motion, and posture. CO/SV measurements are based on a measurement technique known as transbrachial electrovelocimetry (TBEV), which is described in detail below. TBEV measurements yield two time-dependent waveforms: Zo, which represents the basal impedance in the brachial region and is sensitive to slowly changing properties such as blood volume; and ΔZ(t), which characterizes the heartbeat-induced pulse, which changes in contour as blood flows through the brachial region during both systole and diastole. These waveforms are measured from the patient's brachial region, an area somewhat immune to the effects of lung ventilation, which can complicate conventional ICG measurements obtained from the thorax. An algorithm running on a microprocessor within the body-worn monitor analyzes the characteristics of both Zo and ΔZ(t) to determine the value of each TBEV measurement. More specifically, to determine SV/CO/CP values, the monitor relies on "hybrid measurements," which focus on a combination of time-dependent PPG, ECG, and TBEV waveforms measured by a body-worn monitor, along with physiological parameters (e.g., blood pressure) extracted from these waveforms. From these waveforms, parameters such as LVET and PEP can be estimated and used in mathematical relationships to continuously and accurately estimate SV/CO/CP values, as described in detail below. Once measured, they are combined with conventional vital signs and wirelessly transmitted by the body-worn monitor to a central station for efficient patient monitoring.

The TBEV waveform is measured by a small module that is connected to a first set of adhesive electrodes worn in the clavicle/brachial (CB) area of the patient. This area extends roughly from the area near the tip of the shoulder (adjacent the axilla) to the elbow (adjacent the antecubital fossa). The ECG waveform is measured by a small module that is connected to a second set of adhesive electrodes worn on the patient's chest, typically in a conventional Einthoven triangle configuration. Both the TBEV module and the ECG module also include 3-axis accelerometers that measure motion-sensitive acceleration waveforms (ACC). These two accelerometers measure, for example, respiratory-induced chest wall movement, which can be processed to estimate RR; and larger-scale motion, which can be processed to determine motion-related properties such as activity level, posture, motion degree/amplitude, and motion frequency.

During a measurement, the TBEV and ECG modules transmit waveforms and digital information to a wrist-worn transceiver via a wired or wireless connection. The transceiver is also connected to an optical sensor worn on the patient's thumb that measures PPG waveforms generated by an optical system featuring red (approximately 600 nm) and infrared (approximately 900 nm) light-emitting diodes (LEDs). These waveforms can be processed to determine SpO2 values. The wrist-worn transceiver also includes an internal accelerometer that measures an ACC waveform associated with hand movement. These two PPG waveforms, along with the ECG waveform, can be processed to determine cNIBP values.

The TBEV component of the hybrid measurement is measured by injecting a high frequency, low ampere alternating current (AC) field along the course of the brachial artery in the CB region, followed by simultaneous sensing and signal processing of the voltage changes produced within the field. The underlying rational number for TBEV is derived from direct proportion and a high correlation is observed between peak ascending aortic and peak brachial blood flow accelerations. This technique is in stark contrast to the generally accepted volumetric doctrine, proposing that it is the velocity-induced peak rate of change of the specific electrical resistance of axially oriented flowing blood that causes the time-dependent change in the measured impedance. Computationally, the SV of the TBEV determination is obtained by taking the electrical impedance pulse change divided by the square root of the peak rate of change of the base impedance measured in the CB region (i.e., [(dZ/dt) _max /Zo] ^0.5 ). This parameter is then multiplied by LVET and the constant _Vc to obtain SV.

Body-worn monitors also provide technology for measuring cNIBP based on PAT, pulse transit time (PTT), or vascular transit time (VTT), as described in more detail in the patent applications referenced above. These documents describe cNIBP measurements using a "composite approach" (described in detail below) that features several improvements over conventional PAT and PTT measurements.

After the measurement is completed, the body-worn monitor wirelessly transmits the waveforms, vital signs, and SV/CO/CP values to a remote monitor (e.g., a personal computer (PC), a workbench at a nursing station, a tablet computer, a personal digital assistant (PDA), or a cell phone). Typically, the wireless transmitter is located within a wrist-worn transceiver, which also displays and further analyzes this information. Additionally, both the remote monitor and the wrist-worn transceiver may include a barcode scanner, a touch screen display, a camera, a voice and speaker system, and wireless systems that operate over both local area networks (e.g., 802.11 or "WiFi" networks) and wide area networks (e.g., the Sprint network).

In one aspect, for example, the present invention provides a system for measuring both SV and CO from a patient. The system features an impedance sensor connected to at least two patient-worn electrodes and an impedance circuit that processes signals from the at least two electrodes to measure an impedance signal from the patient. An optical sensor within the system is connected to an optical probe and features an optical circuit that measures at least one optical signal from the patient. A body-worn processing system is operably connected to both the impedance sensor and the optical sensor and receives and processes the impedance signal to determine a first value for SV and CO. The body-worn processing system then receives and processes the optical signal to determine a second value for these parameters. Finally, the processing system collectively processes both the first and second values for SV and CO to determine a third value for these parameters, which the processing system then reports to a display device.

In another aspect, the present invention provides a similar system further featuring an ECG sensor connected to at least two body-worn electrodes and featuring ECG circuitry. The ECG circuitry is configured to process signals from the electrodes to measure an ECG waveform and HR values. A processing system is connected to the impedance, optical, and ECG sensors and receives time-dependent waveforms from each of these systems. The processing system then collectively processes the ECG and optical signals to determine blood pressure values, which are then processed to estimate SV and CO.

On the other hand, the present invention provides a similar system featuring ECG, impedance, and optical sensors. Collectively, these sensors generate signals that are processed to determine a collection of SV "estimates." These estimates are then processed using a variety of algorithms to estimate stroke volume.

In another aspect, the present invention provides a method for determining SV, the method comprising the following steps: (a) measuring an impedance signal by an impedance sensor, the impedance sensor being operably connected to a body-worn monitor; (b) measuring an optical signal by an optical sensor; (c) processing the impedance signal to determine a value of (dZ/dt) _maximum ; (d) processing the optical signal to determine a value of SFT; and (e) collectively processing Z ₀ , (dZ/dt) _maximum , and SFT to determine SV.

In another aspect, the present invention provides a method for measuring cardiac output, which is the product of CO and MAP, as described in detail below. Here, CO is determined by processing the ECG waveform to determine the heart rate value, and impedance and optical waveforms are combined to determine SV. MAP is calculated from PAT values measured from PPG and ECG waveforms. Alternatively, MAP is calculated from VTT values measured from TBEV and PPG waveforms. In either case, the combined method processes PAT or PTT to determine MAP.

On the other hand, the present invention provides a body-worn system for measuring SV values from a patient. The body-worn system is characterized by a TBEV module, which includes a circuit configured to inject current adjacent to the patient's arm. The bottom of the circuit includes a pair of electrical connectors, which are configured to snap into a pair of mating connectors arranged on a first electrode, wherein the first connector is configured to inject current into a first portion of the electrode and the second connector is configured to measure a voltage-related signal from a second portion of the electrode. The analog circuit then processes the signal from the second connector to generate a voltage value. A processor connected to the analog circuit converts the voltage value (or a value calculated therefrom) into a time-dependent resistance value, and then converts the time-dependent resistance value into an SV value.

On the other hand, the present invention provides a method for determining SV from a patient, the method comprising the following steps: (a) measuring a motion waveform via a first motion sensor; (b) processing the motion waveform using a motion algorithm to determine motion-related parameters; (c) comparing the motion-related parameters with predetermined threshold parameters to determine whether the patient's motion exceeds an acceptable level; (d) if the patient's motion does not exceed an acceptable level, measuring an impedance waveform from the patient via an impedance sensor; and (e) if the patient's motion does not exceed an acceptable level, calculating a SV value from the impedance waveform.

A composite method for cNIBP is described in detail in the following patent application, which is incorporated herein by reference in its entirety: U.S.S.N. 12/650,354, filed November 15, 2009: BODY-WORN SYSTEM FOR MEASURING CONTINUOUS NON-INVASIVE BLOOD PRESSURE (cNIBP). This composite method incorporates both pressure-dependent and pressure-free measurements and is based on the discovery that the PAT and PPG waveforms used to determine cNIBP are strongly modulated by applied pressure. During a pressure-dependent measurement (also referred to herein as an "indexing measurement"), as pressure is gradually increased toward the patient's systolic blood pressure, two events occur: 1) Once the applied pressure exceeds the diastolic blood pressure, the PAT increases, typically in a nonlinear manner; and 2) as the applied pressure approaches the systolic blood pressure, the magnitude of the PPG amplitude decreases systematically, typically in a linear manner. Applied pressure gradually reduces blood flow and, consequently, blood pressure in the patient's arm, thereby inducing a pressure-dependent increase in PAT. Each of the resulting PAT/blood pressure reading pairs measured during the application of pressure can be used as a calibration point. Furthermore, when the applied pressure equals the SYS, the amplitude of the PPG waveform completely disappears, and PAT is no longer measurable. Using techniques borrowed from conventional oscillometric techniques, the amplitudes of both the PAT and PPG waveforms within appropriate ranges, along with the pressure waveform, are analyzed to derive the patient's SYS, DIA, and MAP, as well as a patient-specific slope related to PAT and MAP. From these parameters, the patient's cNIBP can be determined without the use of a conventional cuff.

The combination of several algorithmic features improves the efficacy of the composite method over conventional PAT measurements of cNIBP. For example, sophisticated, real-time digital filtering removes high-frequency noise from the PPG waveform, allowing its onset to be accurately detected. When processed together with the ECG waveform, this ensures accurate measurement of PAT and, ultimately, cNIBP values. A pressure-dependent indexing method (performed during inflation of an arm-worn cuff) yields multiple data points related to PAT and blood pressure during a short (approximately 60-second) measurement period. Processing of these data points yields an accurate, patient-specific slope that correlates PAT with cNIBP. The inclusion of multiple accelerometers yields a variety of signals that can measure characteristics such as arm height, motion, activity level, and posture, which can be further processed to improve the accuracy of cNIBP calculations and, moreover, allow them to be performed in the presence of motion artifacts. And a model based on femoral arterial blood pressure (which is more representative of the pressure in the patient's core) reduces effects such as "pulse pressure amplification," which can elevate blood pressure measured at the patient's extremities.

The composite method can also include an "intermediate" pressure-dependent measurement, in which the cuff is partially inflated. This partially reduces the amplitude of the PPG waveform in a time-dependent manner. This pressure-dependent reduction in amplitude can then be "fitted" with a numerical function to estimate the pressure at which the amplitude completely disappears, representing systolic pressure.

For pressure-dependent measurements, a small pneumatic system attached to the cuff inflates the bladder so that pressure is applied to the underlying artery according to a pressure waveform. The cuff is typically placed on the patient's upper arm, adjacent to the brachial artery, and the time-dependent pressure is measured by an internal pressure sensor within the pneumatic system (such as an in-line Wheatstone bridge or strain gauge). The pressure waveform gradually ramps up in a predominantly linear manner during inflation, followed by rapid deflation via an "exhaust valve" during deflation. During inflation, as the applied pressure approaches the DIA, mechanical pulsations corresponding to the patient's heartbeat are coupled into the bladder. The mechanical pulsations modulate the pressure waveform so that it comprises a series of time-dependent oscillations. The oscillations are similar to those measured oscillometrically by an automated blood pressure cuff, except that they are measured during inflation rather than deflation. The oscillations are processed as described below to determine a "processed pressure waveform" (thereby directly determining MAP) and indirectly determining SYS and DIA.

Pressure-dependent measurements performed during inflation have several advantages over conventional, similar measurements performed during deflation. For example, inflation-based measurements are relatively fast and comfortable compared to those performed during deflation. Most conventional cuff-based systems using deflation-based oscillometrics take approximately four times longer to perform pressure-dependent measurements than the composite method. Inflation-based measurements are possible due to the relatively slow inflation rate of the composite method (typically 5 to 10 mmHg/sec) and the high sensitivity of the pressure sensors used in the body-worn monitor. In addition, measurements performed during inflation can be terminated immediately once the systolic pressure is calculated. In contrast, conventional cuff-based measurements performed during deflation typically apply pressures far exceeding the patient's systolic pressure; the pressure within the cuff is then slowly vented to below the DIA in order to complete the measurement.

The pressure-free measurement is followed by a pressure-dependent measurement and is typically performed by measuring PAT using the same optical and electrical sensors used in the pressure-dependent measurement. Specifically, the body-worn monitor processes the PAT and other properties of the PPG waveform, along with the patient-specific slope and SYS, DIA, and MAP measurements obtained during the pressure-dependent measurement, to determine cNIBP.

The present invention, in general (and in particular for hybrid measurements of SV/CO/CP), features numerous advantages over conventional techniques for measuring these properties. Compared to TDCO and Fick, for example, body-worn monitors facilitate continuous, noninvasive measurement of these values that is highly accurate and carries a low risk of harmful complications (such as infection and pulmonary artery catheter perforation). And unlike TDCO, Fick, and Doppler-based measurements, hybrid measurements do not require a specially trained observer. TBEV measurements are performed at the brachial level, which inherently offers several advantages over conventional ICG measurements taken from the thorax. For example, complications in the pulmonary system (i.e., intrapleural fluid and pulmonary edema) do not affect SV/CO/CP values measured from this region. Similarly, the baseline transbrachial quasi-static impedance, Zo, is unaffected by medical devices sometimes present in the thorax (such as chest tubes, external pacemaker leads, and central venous catheters). Typically, thoracic ICG measurements require eight separate electrodes, while the TBEV measurements described herein require only two. Finally, in the absence of lung ventilation and the influence of pulmonary artery pulsation, the signal-to-noise ratio of the waveform measured from the brachial artery is relatively high.

These and other advantages of the invention will be apparent from the following detailed description and from the claims.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 shows a schematic diagram of an algorithm for performing hybrid measurement for determining SV/CO/CP values according to the present invention;

FIG2 shows a schematic diagram of a custom electrode patch used to perform the TBEV measurements described in FIG1 ;

FIG3 shows a schematic diagram of the human body's circulatory system indicating the regions where both conventional ICG and TBEV measurements are performed;

FIG4 shows a schematic diagram illustrating the positioning of a body-worn monitor of the present invention relative to both the aorta and the brachial artery and how the diameters of these vessels change during systole and diastole;

FIG5 shows a schematic diagram of a body-worn monitor performing the hybrid measurement shown in FIG1 ;

FIG6A shows a schematic diagram of an alternative embodiment of a body-worn monitor performing the hybrid measurement shown in FIG1 ;

FIG6B shows a three-dimensional diagram of a TBEV module and custom electrodes for use in the alternative embodiment shown in FIG6A ;

FIG7 shows a schematic diagram of analog and digital circuits for performing TBEV measurements according to the present invention;

FIG8 shows a flow chart of an algorithm for calculating SV in which systolic flow time (SFT) is determined using Weissler regression;

FIG9 shows a flow chart of an algorithm for calculating SV using the SV estimator shown in FIG1 ;

FIG10 shows a three-dimensional image of a wrist-worn transceiver that is part of the body-worn monitor shown in FIG5 and FIG6A;

11A to 11E are graphs showing time dependencies of ECG, ICG, TBEV, d(ICG)/dt, and d(TBEV)/dt waveforms, respectively, measured using the body-worn monitor of FIG. 5 ;

12A-12B show graphs of the time dependence of dZ/dt waveforms measured from the thorax and brachia, respectively;

13A to 13D illustrate time-dependent graphs of: a TBEV waveform; a derivative of the TBEV waveform showing a reference point near a pulse maximum for determining pulse onset; a derivative of the TBEV waveform showing a shaded region where Weissler regression is used to estimate LVET; and a derivative of the TBEV waveform where SFT is determined, respectively.

FIG14 shows a correlation plot comparing SFT measured using a “fusion” method that relies in part on HR, pulse rate (PR), and Weissler regression with LVET measured by Doppler/ultrasound;

15A to 15E show two-dimensional Doppler/ultrasound images of the brachial artery measured during systole ( FIG. 15A to FIG. 15C ) and diastole ( FIG. 15D , FIG. 15E );

FIG15F shows a time-dependent graph of the ECG waveform and the TBEV waveform corresponding to the two-dimensional images shown in FIG15A to FIG15E;

16A and 16B are graphs showing time dependencies of waveforms extracted from Doppler/ultrasound images acquired from the humerus similar to those shown in FIG. 15A to FIG. 15E and simultaneously measured TBEV waveforms, respectively;

16C and 16D are graphs showing time-dependent derivatives of the waveforms shown in FIG. 16A and FIG. 16B , respectively;

FIG17 shows derived TBEV waveforms measured from 7 unique subjects;

FIG18A and FIG18B show a correlation diagram and a Bland-Altman plot, respectively, comparing CO measured using Doppler/ultrasound and TBEV in 23 subjects;

19A and 19B are graphs showing time-dependent ICG and TBEV waveforms corresponding to high and low values of CO measured when a subject wore military anti-shock pants (MAST), respectively;

20A to 20C respectively show time-dependent graphs of: an unfiltered TBEV waveform showing both cardiac and respiratory events; a TBEV waveform filtered with a 0.5 Hz to 15 Hz bandpass filter showing only cardiac events; and a TBEV waveform filtered with a 0.001 Hz to 1 Hz bandpass filter showing only respiratory events;

21A and 21B are graphs showing time dependence of TBEV and ACC waveforms modulated by respiratory events, respectively;

21C and 21D show frequency domain power spectra of the time dependence graphs shown in FIG. 21A and FIG. 21B , respectively;

FIG22A shows a graph of the time dependence of TBEV and ECG waveforms measured during a period of no motion;

FIG22B shows a time-dependent plot of the ACC waveform measured simultaneously with the TBEV and ECG waveforms of FIG22A during a period of no motion;

FIG23A shows a graph of the time dependence of TBEV and ECG waveforms measured during exercise;

FIG23B shows a time-dependent graph of the ACC waveform measured simultaneously with the TBEV and ECG waveforms of FIG22A during exercise;

FIG24 shows a schematic diagram of a patient and the overlying axes used with the algorithm and ACC waveform to determine patient posture;

FIG25A shows a graph of time-dependent ACC waveforms measured from a patient's chest during different postures;

FIG25B shows a graph of time-dependent posture determined by processing the ACC waveform of FIG25A using the algorithm and coordinate axes shown in FIG24 ;

FIG26 shows a schematic diagram of ECG and PPG waveforms and how the profiles of the PAT and PPG waveforms determined from these waveforms can be analyzed together to determine an estimate of SV;

FIG27 shows a three-dimensional diagram of a harness for performing TBEV measurements according to an alternative embodiment of the present invention;

FIGURE 28 shows a schematic diagram of an alternative embodiment of the present invention in which the TBEV circuit is contained within a module that is attached directly to the humerus and detached from the ECG circuit, which is worn on the chest;

FIG29 shows a schematic diagram of the TBEV module shown in FIG28 ;

FIG30 shows a schematic diagram of an alternative embodiment of the present invention in which a body-worn monitor wirelessly transmits information between a chest-worn module and a wrist-worn transceiver and from there to a remote monitor; and

31 shows a schematic diagram of a body-worn monitor (similar to that shown in FIG. 30 ) that wirelessly transmits information both between a chest-worn module and a wrist-worn transceiver and to a remote monitor.

Detailed Description of the Invention

Measurement Overview

1 , the present invention features a body-worn monitor that continuously and noninvasively determines SV from TBEV measurements 7 collected from the patient's CB region, along with a series of SV "estimates" 1-6 calculated from cNIBP measurements. Such body-worn monitors are described, for example, in the following patent applications (the contents of which are incorporated herein by reference): U.S.S.N. 12/560,077, filed September 15, 2009, BODY-WORN VITALSIGN MONITOR; and U.S.S.N. 12/762,726, filed April 19, 2009, BODY-WORN VITALSIGN MONITOR. The SV measurement derived using TBEV 7 and the estimates 1-6 can be incorporated into a "hybrid measurement" 10, which operates on a microprocessor within the body-worn monitor and ultimately determines CO and CP.

TBEV is a variant of conventional bioimpedance techniques such as ICG and measures waveforms from the CB region to determine time-dependent parameters such as systolic flow time (SFT), (dZ/dt) _max , and Zo. These parameters are fed into Equation 3 below, where they are coupled with the static parameter _Vc to determine SV.

Here, SV is obtained by taking the square root of the peak rate of change of the transbrachial impedance, Zo, for each TBEV pulse. This parameter is then multiplied by the SFT and the constant amplitude _Vc to yield SV. The derivation of Equation 3 is described in detail in U.S. Patent 6,511,438 and in the following references (the entire contents of which are incorporated herein by reference): Bernstein et al., Stroke Volume Obtained By Electrical Interrogation of the Brachial Artery: Transbrachial Electrical Bioimpedance Velocimetry. Unpublished manuscript, submitted in 2012. Equation 3 assumes that (dZ/dt) _max / _Zo represents the dimensionless acceleration of the blood (in units of 1/ ^s2 ), which is the ohmic analog of the peak aortic blood flow acceleration (cm/ ^s2 ). The forceful systolic ejection from the left ventricle of the heart causes the red blood cells to align parallel during systolic flow so as to produce a pulsatile increase in conductivity. For this model, V _c is estimated purely by weight and does not depend on any factors that depend on the electrode spacing.

Along with SV, the body-worn monitor simultaneously measures cNIBP values (SYS, DIA, MAP, and PP) using a cuffless technique (referred to as the "hybrid approach"), which is described in detail above. According to the hybrid approach, SV is determined explicitly from the TBEV waveform and can be estimated from the cNIBP value. From these parameters, multiple estimators 1-7 are determined, which are collectively processed by an algorithm 10 to determine SV. In addition, the body-worn monitor features multiple accelerometers that generate time-dependent ACC waveforms, which are then further processed by a motion algorithm 11 to estimate the patient's motion level. A function 8 for filtering and rejecting bad data processes information from both estimators 1-7 and the motion algorithm 11 to determine a set of valid data points, which are then linearly combined with another function 9 to determine the final value of SV. Valid data points are, for example, relatively uncontaminated by motion artifacts; they are determined when the motion algorithm 11 compares parameters extracted from the ACC waveform to a predetermined "motion threshold" value. If the parameter exceeds a predetermined threshold, the corresponding SV value is rejected by the function 8. On the other hand, if the parameter is below the predetermined threshold, the corresponding SV value is approved by the function 8 and the SV value is passed to the linear combination algorithm 9, where the SV value is processed to determine the final value of the SV.

Different thresholds may be applied to the SV calculated by TBEV 7 (a measure that is particularly sensitive to motion) and the SV estimated by blood pressure-related estimators 1-6 (which are less sensitive to motion). For example, the motion algorithm 11 may determine that there is a small amount of motion, and therefore the linear combination algorithm 9 relies entirely on the SV value determined by the blood pressure-related estimators 1-6. Alternatively, it may determine that there is a large amount of motion, and in response, the linear combination algorithm 9 will not report an SV value. If the motion algorithm 11 determines that there is no motion, the linear combination algorithm 9 will typically report the SV value determined entirely by TBEV 7.

In an embodiment, the linear combination algorithm 9 combines the different estimators using a simple average or a weighted average to determine a single SV value. More complex methods can also be used to process the estimators. For example, a specific estimator can be selected based on the patient's physiological state or biological parameters (e.g., their age, sex, weight, or height).

Once SV is determined, it can be further processed as defined in Equation 1, Equation 2 to determine both CO and CP.

After the measurements are completed, the body-worn monitor wirelessly transmits the SV/CO/CP values, along with routine vital signs, to a remote processing system. For example, this data can flow through a hospital-based wireless network to a central computer connected to an electronic medical record system. From there, medical professionals such as doctors, nurses, and first responders can evaluate a range of physiological values corresponding to the patient to make a diagnosis. Typically, patients wear a body-worn monitor while they are transferred from the ambulance to the hospital and ultimately to their home.

TBEV measurement (described in detail below) injects a low-ampere, high-frequency current into the patient's CB region and monitors the voltage, which is related to the time-dependent resistance encountered by the current, using Ohm's law (V = I × R). This is based on the assumption that the brachial artery (the only major artery in the CB region) experiences little volume expansion during systole and that changes in resistance are therefore entirely due to acceleration-induced alignment of red blood cells within this artery. In other words, as blood flows through the artery with each heartbeat, the diameter of the brachial artery remains relatively constant, but the acceleration of the blood causes the red blood cells to align. This physiological process thus increases conductivity and decreases resistance in the artery. The time-dependent resistance in the artery manifests as a first waveform (called ΔZ(t)), which is characterized by a series of pulses, each corresponding to a unique heartbeat. The second TBEV waveform (Zo) is filtered to reflect only the baseline impedance of the artery and is sensitive to relatively low-frequency processes such as blood volume, interstitial fluid, and (occasionally) respiratory rate.

Estimators used to determine SV from blood pressure values include the Lilijestrand 1, Wesseling 2, MAP 3, and Herd 4 estimators. These estimators depend linearly on blood pressure values and are shown below in Table 1. In this table, SYS _area refers to the area under the PPG waveform during systole, _ΔPn is the beat-to-beat blood pressure change, _Tn is the duration of the cardiac cycle, and _τn is the time constant that controls the internal circulation dynamics of the Windkessel model.

These estimators are summarized in detail in the following references (and elsewhere), the contents of which are incorporated herein by reference: Chen, Cardiac Output Estimation from Arterial Blood Pressure Waveforms using the MIMIC II Database; Master's thesis submitted to the Massachusetts Institute of Technology; (2009); and Parlikar et al., Model-Based Estimation of Cardiac Output and Total Peripheral Resistance; unpublished manuscript, available at http://lcp.mit.edu/pdf/Parlikar07.pdf . Blood pressure-based estimators can be determined using composite methods or by conventional cuff-based methods such as oscillometric measurement or auscultation.

Table 1 – Estimated amounts of CO

Other SV estimators that can be processed by the algorithm include those based on PAT 5 (which are determined using PPG and ECG waveforms measured by body-worn monitors) and are described in the following reference, the contents of which are incorporated herein by reference: Wang et al., The non-invasive and continuous estimation of cardiac output using a photoplethysmogram and electrocardiogram during incremental exercise; Physiol. Meas.; 31:715-726 (2010). Figure 26 and Equation 4 below indicate Wang's method for analyzing PAT, ECG 375, and PPG 380 waveforms to determine relative values of CO.

CO＝D×[C-ln(PAT)}×(1-IPA)×(1+IPA) ^-1 (4)

In equations D and C there are constants defined in the Wang reference, and IPA is schematically shown in FIG26 . This theory assumes that the PPG waveform 380 includes a well-defined dichoric notch, which allows the parameters _Xi , _X2 , and ultimately IPA to be determined. The integration of the area under the PPG before the notch yields _Xi , while the integration of the area under the PPG after the notch yields _X2 . IPA is defined as the ratio of _Xi to _X2 . Once determined, PAT and IPA are used in equation 4 to derive another estimate of SV/CO/CP.

The PPG waveform (by itself) can be analyzed and used as the "other" estimator 6 of algorithm 10. This waveform represents the time-dependent volume expansion of the underlying artery from which it is measured and is therefore different from the traditional cNIBP waveform (such as that measured using an indwelling arterial catheter), which represents the time-dependent pressure in the artery. However, the PPG and cNIBP waveforms share similar morphologies, especially over relatively long time periods, and can be analyzed to estimate both hemodynamics and, therefore, SV. The following reference (the contents of which are incorporated herein by reference) describes analytical methods for processing waveforms to extract these parameters: Lu et al., Continuous cardiac output monitoring in humans by invasive and noninvasive peripheral blood pressure waveform analysis; J Appl Physiol 101:598-608 (2006).

In yet other embodiments, the "other" estimator for SV/CO/CP measurement can be based on a measurement technique performed by an external sensor that is connected to a body-worn monitor. Such connection can be performed using wired or wireless means. For example, SV can be estimated using techniques such as near-infrared spectroscopy (NIRS), as described in the following reference, the contents of which are incorporated herein by reference: Soller et al., Noninvasively determined muscle oxygen saturation is an early indicator of central hypovolemia in humans; J Appl Physiol 104:475-481 (2008). Thus, a sensor incorporating NIRS measurement can be integrated with a body-worn monitor and attached to the patient's body during measurement. The value of SV calculated using this sensor is sent to the monitor via a wired or wireless connection and can be incorporated into algorithm 10 to further improve the accuracy of the continuous, non-invasive measurement of SV. In all cases, the set of estimators 1-6 is related to CO by a calibration factor (k in Table 1) determined from the absolute measurement of SV, which is provided by TBEV measurement 7 in algorithm 10. Typically, TBEV determines SV to within approximately +/- 20%. Perhaps more importantly, estimators 1-6 and TBEV measurement 7 determine SV by completely different methods and from different locations on the body. Therefore, it is possible that combining the measurements into a single algorithm 10 can reduce errors caused by well-known physiological effects that are typically isolated to these locations.

As shown in FIG2 , TBEV measurements are typically performed using a pair of custom electrodes 24, each featuring two conductive regions 45, 47. The outer conductive region 45 of each electrode 24 (i.e., the region furthest from the CB region) injects a low-ampere (<5 mA), high-frequency (50-100 kHz) current into the patient's CB region. The inner conductive region 47 then measures the time-dependent voltage across the artery. As described above, this voltage change is due to the resistance change caused by blood flowing through the brachial artery, and more specifically, to the acceleration-induced alignment of red blood cells that occurs with each heartbeat. This physiology provides the basis for the mathematical model shown above in Equation 3.

Each conductive region 45 and 47 is typically composed of a conductive "liquid gel" material that roughly matches the impedance properties of human skin. The liquid gel is deposited on top of a conductive substrate coated with a large-area Ag:AgCl thin film, which is in turn deposited on top of the flexible substrate 23. The liquid gel can be, for example, a sponge-like material saturated with a conductive gel or fluid. Adjacent conductive regions 45, 47 are electrically isolated from each other and individually connected to a pair of electrical leads 52A, 52B adhered to the flexible substrate 23 via a pair of separate conductive traces 54A, 54B. The electrical leads 52A, 52B can be, for example, metal rivets or posts that easily snap into corresponding female connectors. An insulating adhesive layer (not shown) interspersed between the conductive regions 45 and 47 electrically isolates these portions of the electrode 24 and is coated with an adhesive that allows it to be securely attached to the patient during measurement.

Figures 3 and 4 illustrate some advantages of TBEV measurements (performed in the CB region 58 of patient 20) compared to conventional ICG measurements (performed in the thoracic cavity 59). As clearly illustrated in Figure 3 (which has underlying imagery borrowed from Gray's Anatomy), the thoracic cavity 59 features a large and complex collection of arteries and veins, as well as most of the patient's vital organs (such as their lungs, heart, kidneys, liver, stomach, and gastrointestinal tract). Each of these systems (and more specifically, the lungs and the aorta, which originate from the left side of the heart) contains conductive fluids (e.g., blood and lung fluid) that can affect conventional ICG waveforms. For example, physiological processes such as pulmonary fluid overload, pulmonary edema, and lung injury can alter the time-dependent impedance characteristics of the patient's thoracic cavity, such that the waveform measured thereby will no longer reflect the true hemodynamic state. In stark contrast, the humerus 58 is physically remote from the influence of the lungs and features only one large artery (the brachial artery), which affects TBEV measurements. This ultimately simplifies the morphology of the TBEV waveform and reduces its patient-to-patient variability, thereby simplifying the calculation of SV.

Importantly, previous studies have shown a strong correlation between peak blood flow acceleration in the aorta (where SV first manifests) and peak blood flow acceleration in the brachial artery (where TBEV measures the signal used to estimate SV via square-root transformation). In terms of velocity, peak aortic blood flow velocity is approximately 80-124 cm/s (average approximately 100 cm/s), while peak aortic blood flow velocity in the brachial artery is approximately 30-70 cm/s (average approximately 50 cm/s). Experiments for measuring these parameters are described in the following references, the contents of which are incorporated herein by reference: Gardin JM et al., Evaluation of bloodflow velocity in the ascending aorta and main pulmonary artery of normal subjects by Doppler echocardiograpy. Am. Heart J. 1984; 107: 310; Wilson S et al., Normal intracardiac and great artery blood velocity measurements by pulsed Dopple rechocardiography. Br. Heart J. 1985; 53: 451; Fronek A., Non invasive diagnostics invascular disease. McGraw-Hill, N.Y. 1989, p. 117; Green D et al., Assessment of brachial artery blood flow across the cardiac cycle: retrograde flows during bicycle ergometry. J. Appl. Physiol 2002; 93: 361. These references generally indicate that the average blood flow velocity in the aorta is approximately twice that in the brachial artery.

FIG4 illustrates how the complexity of the TBEV signal is further reduced compared to the ICG signal. Without being bound by any theory, this may be due to the relatively complex time-dependent behavior of the vasculature in the thoracic cavity (e.g., aorta 66A, 66B) compared to the time-dependent behavior of the vasculature in the CB region (e.g., brachial 68A, 68B). More specifically, the figure shows the position of the body-worn monitor 19 on the patient, along with a schematic diagram of the patient's aorta 66A, 66B and brachial arteries 68A, 68B. During systole, the left ventricle contracts to force blood into the aorta 66A at an ejection volume defined as SV. As the cardiac cycle transitions from diastole to systole, this process creates two simultaneous processes in the aorta: 1) an increase in volume along the aortic wall, which is highly elastic and expands to an enlarged state 66A during systole and then contracts to a relaxed state 66B during diastole; and 2) an acceleration-induced alignment of red blood cells within the arterial lumen, causing these cells to move from a random orientation during diastole to an aligned, parallel orientation during systole. Without being bound by any theory, it is likely that both the volume and acceleration-induced alignment processes occur in the aorta during the cardiac cycle. Both processes affect the conductivity of the blood in the aorta in a patient-specific manner, thereby complicating the pulsatile component of the ICG signal and making it difficult for a single mathematical equation to characterize a large group of patients. Historically, parameters extracted from the ICG signal were fed into the well-known Sramek-Bernstein equation (shown below as Equation 5), which is based on a volume expansion model:

In Equation 5, δ represents compensation for body mass index, Zo is the baseline impedance, and L is estimated from the distance separating the current injection electrode and the voltage measurement electrode on the chest. This equation and several mathematical derivatives are described in detail in the following reference, the contents of which are incorporated herein by reference: Bernstein, Impedance cardiography: Pulsatile blood flow and the biophysical and electrodynamic basis for the stroke volume equations; J Electr Bioimp; 1:2-17 (2010). Equation 5 depends on LVET, which is estimated from each pulse in the ICG waveform, as described in more detail below. Both the Sramek-Bernstein equation and its earlier derivative (called the Kubicek equation) are characterized by a "static component" _Z0 and a "dynamic component" ΔZ(t) (the dynamic component is related to LVET and (dZ/dt) _max / _Zo value), which is calculated from the derivative ΔZ(t) of the raw ICG signal. These equations assume that: (dZ/dt) _max / _Zo represents the radial velocity of the blood due to the volume expansion of the aorta (in Ω/s).

In contrast to the aorta, the brachial artery is a relatively muscular vessel that undergoes little expansion during systole 68A and contraction during diastole 68B; its arterial volume (as shown in FIG4 ) therefore remains relatively constant during the cardiac cycle. Consequently, time-dependent variations in the arterial waveform are almost entirely due to the periodic, sinusoidal, heartbeat-induced parallel alignment of the red blood cells within the artery. Ultimately, this means that, in order to develop a basic mathematical model for the brachial artery, it is not necessary to estimate the relative contributions of volume expansion and red blood cell alignment, which, as described above, can vary from patient to patient.

Sensor Configuration

Referring to Figures 5 and 6 , in a preferred embodiment, a body-worn monitor 19 is distributed across a patient 20 to measure SV/CO/CP. The monitor features a TBEV module 22 (worn near the CB region) attached to the patient 20 using a first two-part electrode 24, shown in Figure 2 . A second two-part electrode 28 is attached near the patient 20's elbow. As described above, the outer conductive region in the first two-part electrode injects a high-frequency, low-ampere current into the patient's CB region, while the outer conductive region in the second two-part electrode acts as a sink for this current. Simultaneously, the inner electrode measures a voltage according to Ohm's law, which describes the resistance encountered by the propagating current. Each of the first two-part electrodes 24 features a rivet or post that snaps into a paired female component in the TBEV module, directly connecting these components to the analog circuitry therein. Similarly, the electrodes in the second two-part electrode 28 connect to snaps embedded in a cable 26 that connects the first and second electrodes 24 and 28. Cable 26 includes conductors for transmitting digital data via the Control Area Network (CAN) protocol. The use of this protocol is described in detail in the following patent application, U.S.S.N. 12/560,077, filed September 15, 2009, entitled "BODY-WORN VITAL SIGN MONITOR," previously incorporated herein by reference. Cable 26 also includes conductors for transmitting analog signals to TBEV module 22 for measuring the aforementioned voltages. An analog-to-digital converter (not shown) within the module digitizes the TBEV waveform to form ΔZ(t) and Zo, which are then analyzed (as described above) by a microprocessor (also not shown) to determine the value of SV.

The ECG module 40, worn on the patient's chest, is connected to the TBEV module 22 via a similar cable 27, which consists solely of conductors for transmitting digital signals according to the CAN protocol. The ECG module 40 is connected to a set of three disposable ECG electrodes 42A-C, which are arranged on the patient's chest in a conventional "Einthoven triangle" configuration via a corresponding set of three ECG leads 44A-C. During a measurement, the ECG module 40 measures the analog signals from each electrode 42A-C and lead 44A-C and performs differential amplification of these signals according to techniques known in the art to generate an ECG waveform. An analog-to-digital converter (not shown) digitizes the ECG waveform, and a microprocessor (also not shown) analyzes the well-known QRS complex within this waveform using a beat-picking algorithm to determine the HR value. The digital representation of this data is transmitted via cable 27 in a CAN-formatted packet to a CAN transceiver (not shown) within the TBEV module 22. There, the packets are combined with corresponding packets containing the TBEV waveform and SV value, calculated as described above. These packets travel through the CAN conductors in cable 26, pass through a second two-part electrode 28, and then via a third cable 29 to a wrist-worn transceiver 30, which is attached to the patient's wrist using a plastic bracket 32 and Velcro strap 34. These components are described in more detail in the following co-pending patent applications, the contents of which have been previously incorporated herein by reference: U.S.S.N. 12/560,077, filed September 15, 2009, BODY-WORN VITAL SIGNMONITOR; and U.S.S.N. 12/762,726, filed April 19, 2009, BODY-WORN VITAL SIGNMONITOR. Additionally, wrist-worn transceiver 30 is connected via a short cable 38 that carries only the analog signal measured by thumb-worn optical sensor 36. Within the wrist-worn transceiver is a pulse oximetry circuit (not shown) that converts the signals measured by the optical sensor 36 to generate a PPG waveform and a corresponding value for SpO2. A microprocessor within the wrist-worn transceiver 30 processes the PPG and ECG waveforms to generate PAT values, or processes the TBEV and PPG waveforms to generate VTT values. These transit times are converted to cNIBP values using the composite method described above. The cNIBP values are, in turn, converted to an estimate of SV using the algorithm shown in FIG1 . Thus, the corresponding value for CO is determined using the SV and HR values measured by ECG, while the value for CP is determined using MAP and CO.

Technically, TBEV-based measurement of SV requires only an isolated TBEV waveform and can be performed in the absence of an ECG waveform. However, this signal, which is relatively easy to measure and represents the start of the cardiac cycle associated with each heartbeat, can be used to "gate" the relatively weak TBEV signal to make it easier to extract the properties described above in Equation 3. More specifically, the software beat picker described above can detect QRS complexes within the ECG waveform, which are associated with the start of individual heartbeats. The relevant portion of the TBEV waveform is typically several hundred milliseconds after the QRS complex. Analysis of these portions of the TBEV waveform yields properties such as SFT, (dZ/dt) _maximum , and Zo, which are used to calculate SV as described above. Gating the TBEV waveform in this manner can be particularly effective in analyzing these properties when noise is present in the TBEV waveform (e.g., during exercise).

The HR determined from the ECG waveform is used to convert SV to CO and CO to CP, as described above in Equations 1 and 2. Typically, HR is determined from the time period separating adjacent QRS complexes in the ECG waveform; alternatively, it can be estimated from adjacent pulses in a PPG or TBEV waveform.

Additionally, the ECG module 40 can be connected to either 5 or 12 leads. It is typically hardwired into the TBEV module 22. The third cable 29 plugs into the wrist-worn transceiver 30 using a detachable connector 31, which allows it to be easily removed. In other embodiments, the order of the ECG module 40 and the TBEV module 22 can be reversed, such that the TBEV module 22 is closer to the chest and the ECG module 40 is closer to the CB area. In still other embodiments, the TBEV module 22 can be placed on the third cable 29 and attached directly to the second electrode 28, while the ECG module 40 can be placed in its original location on the TBEV module 22 and enclosed by a housing that connects to the first electrode 24 and feeds the analog signal into the TBEV module 22. In general, multiple configurations of the various modules, cables, and electrodes shown in FIG. 6 are within the scope of the present invention.

An alternative embodiment of the present invention is shown in Figures 6A and 6B. Here, the ECG module 40 is worn on the patient's CB area, while the TBEV module 22 is worn near the elbow. Both the ECG 40 and TBEV 22 modules are attached to the patient via two-part electrodes 24, 28, with the TBEV electrode 24 shown in more detail in Figure 6B. The two-part electrode 24 features a pair of female snaps 62A, 62B arranged on a flexible substrate 64, which is connected to underlying conductive areas 63A, 63B. Each conductive area 63A, 63B features a solid gel material selected to match the electrical impedance characteristics of human skin, deposited on an Ag/AgCl film. The flexible substrate 64 features an underlying adhesive layer that securely attaches the electrode 24 to the patient's CB area during use. Female snaps 62A, 62B are selected to geometrically match a pair of metal rivets 61A, 61B attached to the bottom of the TBEV circuit board 60. During use, the rivets 61A, 61B snap into the female snaps 62A, 62B, securing the TBEV module 22 to the patient. The rivet 61B furthest from the CB region connects to the TBEV circuit and injects current through the corresponding conductive region 63B, while the rivet 61A closest to the region measures the corresponding voltage. A plastic housing 65 covers the TBEV circuit board 60 and protects it from liquids and other materials present in the hospital. Note that in the figure, the female snaps 62A, 62B are positioned on the electrodes 28, while the metal rivets 61A, 61B are positioned on the circuit board 60. However, in an alternative embodiment, these components may be reversed, ie, the female snaps 62A, 62B may be disposed on the bottom of the circuit board and the metal rivets 61A, 61B may be disposed on top of the electrodes 28 .

The ECG module 40 is attached to the second electrode 24 in a geometry similar to that shown in FIG6B . However, here, the electrode 24 is not electrically connected to the internal ECG circuitry, but rather serves solely to hold the ECG module 40 in place. The conductive areas of the electrodes are connected via electrical leads within the module 40, which are then connected to the TBEV module 22 via cable 68, which is not connected to the ECG circuitry. There, the module 22 receives and centrally processes the signals from the first and second electrodes 28, 24 to measure the time-dependent voltage as described above. This voltage is then converted into a time-dependent resistance to form a TBEV waveform, which is then processed to determine SV and ultimately CO. Cable 68 also includes conductors for transmitting packets containing the digitized ECG waveform and HR according to the CAN protocol. These packets pass through the CAN transceiver in the TBEV module 22, through the third cable 29, and ultimately to the wrist-worn transceiver 30 for further processing and display.

In FIG6A , individual ECG electrodes 42A-C are arranged on a single chest-worn patch 67 that is attached to the patient near the middle of the chest. Patch 67 is connected to ECG module 40 via a single cable 69 that includes a separate conductor corresponding to each electrode 42A-C. These conductors transmit analog signals to ECG module 40, where they are analyzed to determine the ECG waveform and HR as described above.

Within the body-worn monitor 19 are three triaxial accelerometers that measure ACC waveforms corresponding to the x-, y-, and z-axes. The accelerometers (not shown) are located within the ECG module 40, the TBEV module 22, and the wrist-worn transceiver 30. During measurement, the ACC waveforms generated by the accelerometers are processed by a microprocessor within each of these components to determine motion-related parameters. If the parameter, or a secondary parameter derived therefrom, falls below a predetermined threshold, the measurement of SV (or any vital sign, for that matter) is rejected. The algorithm used for this calculation is described, for example, in the following co-pending patent application, the contents of which are incorporated herein by reference: VITAL SIGN MONITORING SYSTEM FEATURING 3 ACCELEROMETERS, U.S.S.N. 12/469,094 (filed May 20, 2009).

Within the TBEV module is an analog circuit 100 (shown in FIG7 ) that performs TBEV measurements in accordance with the present invention. The figure shows only one embodiment of the circuit 100; similar electrical results can be achieved using designs and collections of electrical components different from those shown in the figure.

Circuit 100 features a first electrode 105B that injects a high-frequency, low-ampere current ( _Iin ) into the patient's brachial artery. This serves as a current source. Current pump 102 typically provides a modulated current with a modulation frequency typically between 50 kHz and 100 kHz and a current amplitude between 0.1 mA and 10 mA. Preferably, current pump 102 supplies a current with an amplitude of 4 mA, which is modulated at 70 kHz through first electrode 105B. Second electrode 104A serves as a current sink ( _Iout ).

A pair of electrodes 104B, 105A measures the time-dependent voltage encountered by the propagating current. These electrodes are represented in the figure as V+ and V-. As described above, using Ohm's law (V=I×R), the measured voltage divided by the amplitude of the injected current yields the time-dependent resistance to ac (i.e., impedance), which is related to blood flow in the brachial artery. As shown by waveform 128 in the figure, the time-dependent resistance is characterized by a slowly varying dc offset (characterized by Zo) that indicates the baseline impedance encountered by the injected current; for TBEV, this will depend on, for example, the amount of fat, bone, muscle, and blood volume in a given patient's brachial region. Zo (whose value is typically between approximately 10Ω and approximately 150Ω) is also affected by low-frequency, time-dependent processes such as respiration. Such processes affect the TBEV measurement and are manifested in the inherent capacitance near the brachial region by low-frequency fluctuations (such as those shown in waveform 128). A relatively small (typically 0.1-0.5Ω) ac component, ΔZ(t), lies above Zo and is attributed to the resistance changes caused by the propagation of blood in the brachial artery induced by the heartbeat, as described in detail above. ΔZ(t) is processed by a high-pass filter to form the TBEV signal, which is characterized by an aggregation of individual pulses 130, which are ultimately processed to ultimately determine stroke volume and cardiac output.

The voltage signals measured by first electrode 104B (V+) and second electrode 105A (V-) are fed into a differential amplifier 107 to form a single differential voltage signal, which is modulated according to the modulation frequency of current pump 102 (e.g., 70 kHz). From there, the signal flows to a demodulator 106, which also receives the carrier frequency from current pump 102 to selectively extract only the signal components corresponding to the TBEV measurement. The combined functionality of differential amplifier 107 and demodulator 106 can be implemented using a variety of circuits designed to extract weak signals, such as the TBEV signal, from noise. For example, these components can be combined to form a "lock-in amplifier," which selectively amplifies signal components occurring at a well-defined carrier frequency. Alternatively, the signal and carrier frequency can be deconvolved using circuitry, typically featuring one or more diodes, in a manner similar to that used in conventional AM broadcasting. The phase of the demodulated signal can also be adjusted during the amplification process using a phase adjustment component 108. In one embodiment, the ADS1298 chipset series marketed by Texas Instruments can be used for this application. This chipset features a fully integrated analog front end for both ECG and impedance pneumographography. The impedance pneumographography measurement is performed using components for digital differential amplification, demodulation, and phase adjustment (such as those used for TBEV measurement) that are directly integrated into the chipset.

Once the TBEV signal is extracted, it flows through a series of analog filters 110, 112, and 114 within circuit 100, which remove extraneous noise from the Zo and ΔZ(t) signals. A first low-pass filter 1010 (30 Hz) removes any high-frequency noise components that could corrupt the signal (e.g., power line components at 60 Hz). The portion of the signal that passes through this filter 110 (representing Zo) is passed directly to a channel in analog-to-digital converter 120. The remainder of the signal is fed into a high-pass filter 112 (0.1 Hz), which passes the high-frequency signal components responsible for the shape of a single TBEV pulse 130. This signal then passes through a final low-pass filter 114 (10 Hz) to further remove any high-frequency noise. Finally, the filtered signal passes through a programmable gain amplifier (PGA) 116, which amplifies the resulting signal with a computer-controlled gain using a 1.65 V reference. The amplified signal represents ΔZ(t) and is transmitted to a separate channel of an analog-to-digital converter 120, where it is digitized along with Zo. The analog-to-digital converter and PGA are directly integrated into the aforementioned ADS1298 chipset. The chipset can simultaneously digitize waveforms such as Zo and ΔZ(t) at a 24-bit resolution and sampling rate suitable for physiological waveforms (e.g., 500 Hz). Thus, in theory, this single chipset can perform the functions of the differential amplifier 107, the demodulator 108, the PGA 116, and the analog-to-digital converter 120. Relying solely on a single chipset to perform these multiple functions ultimately reduces both the size and power consumption of the TBEV circuit 100.

The digitized Zo and ΔZ(t) waveforms are received by the microprocessor 124 via a conventional digital interface (e.g., an SPI or I2C interface). Algorithms for converting the waveforms into actual measurements of SV and CO are executed by the microprocessor 124. The microprocessor 124 also receives digital motion-related waveforms from the onboard accelerometer and processes them to determine parameters such as motion degree/amplitude, motion frequency, posture, and activity level.

FIG8 and FIG9 illustrate flow charts of algorithms 133A and 133B, which are implemented using compiled computer code, for example, operating on the microprocessor 124 shown in FIG7 . The compiled computer code is loaded into a memory associated with the microprocessor and executed each time a TBEV measurement is converted into CO and SV values. The microprocessor typically runs an embedded real-time operating system. The compiled computer code is typically written in a language such as C, C++, or assembly language. Each step 135-150 in the various algorithms 133A and 133B is typically performed by a function or calculation included in the compiled computer code.

Wrist-worn transceiver

FIG10 shows in greater detail a wrist-worn transceiver 272 used to perform the hybrid SV measurement method according to the present invention. It features an embedded microprocessor (not shown) for performing these calculations and a touchscreen interface 273 for displaying CO/SV and the other features described above. A flexible wristband 290 secures the transceiver 272 to the patient's wrist like a conventional wristwatch. Connected to the transceiver 272 is an analog cable 292 that terminates in an optical sensor (not shown) that wraps around the base of the patient's thumb to measure PPG waveforms. During measurement, the optical sensor generates a series of time-dependent PPG waveforms (measured in both red and infrared wavelengths), which the microprocessor in the transceiver processes (along with TBEV and ECG) to measure cNIBP, SpO2, and provide waveforms for hybrid SV/CO/CP measurements.

As described above, the wrist-worn transceiver attaches to the patient's wrist using a flexible strap 290 that extends through two D-ring openings in the plastic housing 206. Transceiver 272 features a touchscreen display 220 that provides a GUI 273 that changes depending on the viewer (typically a patient or medical professional). Specifically, transceiver 272 includes a small infrared barcode scanner 202 that, during use, scans a barcode on a badge worn by a medical professional. The barcode indicates to the transceiver's software that a nurse or doctor, for example, is viewing the user interface. In response, GUI 273 displays vital sign data and other medical diagnostic information appropriate for the medical professional. Using this GUI 273, a nurse, doctor, or medical professional can, for example, view vital sign information, set alarm parameters, and enter information about the patient (e.g., their demographics, medications, or medical conditions). For example, for the aforementioned SV/CO/CP measurements, the clinician can enter the patient's gender, height, weight, and age. These parameters can be used in the calculations described above in Equation 3 to estimate Vc used in Equation 3 to calculate SV. Once entered, the clinician can press a button on the GUI 273 to indicate that these operations are complete and the appropriate data for SV measurement has been entered. At this point, the display 220 provides an interface more suitable for the patient, such as one that only displays the time of day and battery level.

Transceiver 272 features three CAN connectors 204a through 204c on one side of its upper portion, each supporting the CAN protocol and wiring diagram and forwarding digitized data to the transceiver's internal CPU. Digital signals passing through the CAN connectors include a header that indicates the specific signal (e.g., TBEV, ECG, ACC, or a value calculated from these waveforms) and the sensor from which the signal originated. In an alternative embodiment, some of these digital signals are sent via Bluetooth from a chest-worn module, which maintains the CAN structure of the packet. This allows the CPU to easily interpret signals arriving through CAN connectors 204a through 204c (such as those corresponding to TBEV and ECG waveforms described above) and means that these connectors are not associated with a specific cable. Any cable connected to the transceiver can be plugged into any of connectors 204a through 204c.

As shown in FIG10 , one embodiment of the present invention features a first connector 204a that receives a Bluetooth "adapter" 208. The adapter features an embedded antenna 213 and a connector 209 that snaps into and mates with any of the CAN connectors 204a through 204c. During operation, the adapter 208 automatically pairs with the Bluetooth transceiver in the chest-worn sensor and then receives a digital data stream via Bluetooth through antenna 213. The internal CAN transceiver formats the data stream into CAN-compatible packets and then passes them through connector 209 to the wrist-worn transceiver, where they are processed as described above.

The second CAN connector 104b receives a cable 286 connected to another sensor (e.g., a pneumatic cuff-based system for measuring blood pressure values used in the combined method). This connector 204b receives the time-dependent pressure waveform delivered to the patient's arm by the pneumatic system 285, along with the values of SYS, DIA, and MAP measured during the combined method's index measurement. Once the index measurement is complete, the cable 286 is unplugged from the connector 204b and replugged approximately four hours later for another index measurement.

The final CAN connector 204c may be used for auxiliary devices such as blood glucose meters, infusion pumps, body-worn insulin pumps, NIRS systems, ventilators, or et-CO2 measurement systems. As described above, the digital information generated by these systems will include a header indicating their origin so that the CPU can process them accordingly.

Transceiver 272 includes a speaker 201 that allows a medical professional to communicate with a patient using Voice over Internet Protocol (VOIP). For example, using speaker 201, a medical professional can question a patient from a central nursing station or a mobile phone connected to the hospital's internal wireless, Internet-based network. Alternatively, a medical professional can wear a separate transceiver similar to the one shown and use it as a communication device. In this application, the patient-worn transceiver 272 functions much like a conventional cellular phone or "walkie-talkie": it can be used for voice communication with the medical professional and can also relay information describing the patient's vital signs and movements. The speaker can also announce preprogrammed messages to the patient, such as those used to calibrate a chest-worn accelerometer for posture calculations, as described above.

Clinical data from CO/SV measurements

Figures 11A to 11E show examples of ECG, ICG, and TBEV waveforms, along with time-dependent derivatives of both the ICG (d(ICG)/dt) waveform and the TBEV (d(TBEV)/dt) waveform. These data were measured simultaneously from a human subject over a 5-second period using a body-worn monitor similar to that shown in Figure 5. Each waveform is characterized by a heartbeat-induced "pulse" that indicates a unique physiological process. For example, the ECG waveform shown in Figure 11A was measured using the conventional ECG electrodes and circuitry described above. It is characterized by a conventional QRS complex that indicates rapid depolarization of the right and left ventricles of the heart. Informally, the ECG waveform represents the onset of the cardiac cycle. The ICG waveform (Figure 11B) was measured from the chest using a conventional ICG monitor and chest-worn electrodes configured as described above, and its derived form (Figure 11D) yields the parameters used above in Equation 5 to calculate SV. Specifically, from the derived waveform, the computer program calculates parameters such as (dZ/dt) _max and SFT, which, as described above, correspond to the time period from the opening of the aortic valve (heralding the start of ejection) to the closure of the aortic valve (marking the end of ejection). The TBEV waveform (FIG. 11C) and its time-dependent derivative (FIG. 11E) were measured from the brachial artery using a configuration of arm-worn electrodes similar to that shown in FIG5. The derived TBEV waveform (like the derived ICG waveform) yields similar impedance parameters such as (dZ/dt) _max and SFT, and is characterized by a signal-to-noise ratio similar to that shown by the ICG waveform, despite the fact that its origin is the much smaller brachial artery.

Figures 12A and 12B illustrate how to extract LVET and SFT from both derived ICG and TBEV waveforms, respectively. As shown in Figure 12A, the derived ICG waveform is characterized by continuous pulses, each characterized by three points: point "B" when the pulse rebounds, indicating the opening of the aortic valve; point X when the pulse is at its lowest point, indicating the closing of the aortic valve; and point "C" when it is at its maximum value, indicating the maximum slope of ΔZ(t) when the pulse rebounds, which is equivalent to _{the maximum value of} (dZ/dt). LVET is typically calculated by the time differential between points B and X. However, due to the delicate nature of these fiducial markers, even low levels of noise in the waveform can make them difficult to measure. Ultimately, such noise adds error to the calculated LVET and the resulting SV.

It is relatively easy to determine the SFT from the derived TBEV waveform shown in Figure 12B. Here, there are no complicated B and X points. The initial pick-up or start of the pulse indicates the beginning of flow, which corresponds to the opening of the aortic valve. And a second zero crossing point after the dZ/dt _maximum (occurring later in time) indicates when the acceleration of the red blood cells temporarily stops. The second zero crossing point represents the end of systolic forward flow, which corresponds to the closure of the valve. Computationally, using a computer algorithm, such a determination of the SFT can be easily performed by following the progress of the pulse and noting the appropriate zero crossing points.

Figure 13 A to Figure 13 D indicate the technology of measuring both the start 90 and the dicrotic notch 91 in the pulse contained in the TBEV waveform. As mentioned above, such fiducial markers can sometimes be covered by baseline noise (for example, caused by motion or low signal level), making them difficult to measure. In the waveform derived in Figure 13 B, a pair of data points 92 near the pulse peak can be selected and consistent with a simple line 96, wherein the point where line 96 intersects with a zero value (shown by dotted line 90 in the figure) indicates the start of the pulse. Once this point is determined, the zero crossing point indicating the dicrotic notch can first be initially estimated using an approximation for SFT, and SFT is the time between the start of the pulse and the dicrotic notch. This is done using an equation (shown below as equation 6) called "Weissler regression", which estimates LVET from HR.

LVET＝-0.0017×HR+0.413 (6)

Weissler regression allows LVET (equivalent to SFT) to be estimated from HR, which is measured from the ECG waveform, or from PR, which is measured from the PPG waveform. The LVET determined by the Weissler relationship is also shown as a vertical line in Figures 13A and 13B. Figure 14 shows a correlation plot of the "fused" SFT determined from HR and PR compared to LVET determined by Doppler/ultrasound for a study of 38 subjects. Here, Doppler/ultrasound represents the gold standard for determining LVET. As should be clear from these data, there is a strong correlation between the two methods (r = 0.8), indicating that Equation 6 is a reasonable way to determine SFT. Therefore, this method can be used together with parameters extracted from the TBEV signal measured at the brachial level to estimate SV.

To further support this, Figures 15A to 15E show Doppler/ultrasound images measured from the brachial artery, and Figure 15F shows the simultaneously measured TBEV and ECG waveforms. These data indicate two important aspects of TBEV measurements. First, the Doppler/ultrasound images confirm that the volume expansion of the brachial artery during a typical cardiac cycle is minimal. The diameter of the artery undergoes almost no measurable change during the cycle, which means that the heartbeat-induced changes in blood conductivity (as measured by TBEV) are primarily due to the acceleration of the blood and the subsequent parallel alignment of the red blood cells. Second, the images also show that the dicrotic notch in the TBEV waveform does correspond to a point in time when the acceleration of the blood is temporarily zero, so the SFT can be accurately calculated from this fiducial marker.

More specifically, the figure shows a Doppler/ultrasound image that indicates that the forward blood flow velocity is zero before systole (Figure 15A), thereby reducing the conductivity and the corresponding amplitude of the TBEV waveform. This point marks the beginning of the TBEV pulse. The opening of the aortic valve induces contraction (Figure 15B) and increases the acceleration of blood and therefore the conductivity in the humerus, resulting in a rapid increase in the amplitude of the TBEV waveform. The closure of the aortic valve (as characterized by SFT) marks the end of systole (Figure 15C), a temporary pause in acceleration and therefore the appearance of the dicrotic notch. During diastole (Figure 15D), flow increases again due to reflected waves and blood remaining in the aorta, which is injected into the humerus and eventually decays until the cycle repeats with a new heartbeat (Figure 15E). This relatively simple physiological function contrasts with the complex, underlying physiological processes occurring in the thorax as described above, which are the basis of ICG-based measurement of SV.

Figures 16A to 16D further illustrate how the TBEV waveform derives SFT in a manner almost identical to that of Doppler/ultrasound deriving LVET, which represents the gold standard for this measurement as described above. Here, Figure 16A illustrates a time-dependent waveform extracted from a collection of two-dimensional Doppler/ultrasound images (such as those shown in Figures 15A to 15E). The waveform indicates time-dependent blood flow velocity, and its derived form is shown in Figure 16C. Shown below these waveforms in Figures 16B and 16D are the TBEV waveforms (Figure 16B) and their derived forms (Figure 16D) measured simultaneously. The dotted lines 97A, 97B, 98A, and 98B in the figure illustrate the pulse onset (measured, for example, as shown in Figure 13) and the dicrotic notch from the two sets of waveforms, respectively. As should be clear from the figure, these points are completely consistent, indicating that the LVET clearly measured by the Doppler/ultrasound waveform is almost identical to the SFT measured by the TBEV waveform.

Another advantage of TBEV waveforms over those measured by conventional ICG is that they experience little patient-to-patient variability, making their computer-based analysis relatively easy. Figure 17 demonstrates this by showing waveforms derived from seven different subjects. Each waveform has roughly the same morphology, and in all cases, the relevant fiducial markers (pulse onset, pulse maximum, zero crossing) are clear. This suggests that a simple computer algorithm can be used to extract the (dZ/dt) _maximum and SFT, which are then used to calculate SV as described above.

The above analysis was used in a formal clinical study to test the accuracy of CO measurements using TBEV and Equation 3 above, compared to CO measurements using Doppler/ultrasound. The correlation plot and Bland-Altman plot are shown in Figures 18A and 18B, respectively. The shaded gray area in the figure indicates the inherent error associated with conventional Doppler/ultrasound measurements, which is approximately +/- 20%. For this study, measurements were performed in a total of 23 subjects (11M, 12W) with an age range of 21 to 80 years, and all but two of these subjects had correlations that fell within the error range of Doppler/ultrasound measurements.

Figures 19A and 19B demonstrate that the TBEV waveform measured from the brachial region may be a better determinant of CO than the ICG waveform measured from the thorax. Here, both waveforms were measured while the subject was wearing a "MAST" suit, a pressurized pants used to force blood from the lower extremities toward the trunk and heart. The MAST suit thus simulates the reversal of hemorrhage, resulting in an increase in CO/SV. As shown by the dotted lines in the figure, LVET is measured by ICG, while SFT is measured using TBEV. The waveforms from these techniques are measured simultaneously from the thorax and brachial region during high and low SV measurements. Increased SV is achieved when the MAST suit forces blood into the thorax. Increases in both LVET and SFT indicate an increase in SV. In the ICG waveform (Figure 19A), only a small increase in LVET is detected. In contrast, in the TBEV waveform (Figure 19B), a large increase in SFT is detected, suggesting that measurements made in this region of the body may be more sensitive to small changes in SV and CO.

Respiration rate measurement by TBEV

TBEV, like techniques such as impedance pneumography, injects a small amount of current into the patient's body and measures the resistance encountered by the current (i.e., impedance) to calculate relevant parameters. During TBEV measurements, heartbeat-induced blood flow produces a pulsating component of ΔZ(t). In addition, changes in capacitance due to respiration can also affect the impedance measured by TBEV. Figures 20A to 20C illustrate this. In Figure 20A, for example, the TBEV waveform without digital filtering shows both a high-frequency cardiac component due to blood flow and low-frequency fluctuations due to respiratory rate. Both of these features can be extracted and analyzed using digital filtering. For example, as shown in Figure 20B, processing the TBEV waveform shown in Figure 20A through a first bandpass filter (0.5 → 15 Hz) removes the respiratory component, leaving only the cardiac component. Similarly, processing the TBEV waveform shown in Figure 20A through a second bandpass filter (0.001 → 1 Hz), as shown in Figure 20C, removes the cardiac component, leaving only the fluctuations due to respiration. In the latter case, the peaks in the waveform can be counted by conventional breath picking algorithms to determine the respiratory rate.

The algorithm used to calculate respiratory rate can be expanded to include processing signals from the accelerometer within the TBEV module. For example, as shown in Figures 21A to 21D, these signals can be collectively processed to accurately determine respiratory rate, even in the presence of motion. Similar technology is described in the following co-pending application, the contents of which are incorporated herein by reference: U.S.S.N. 12/762,874, filed April 14, 2010: BODY-WORN MONITOR FOR MEASURING RESPIRATION RATE. More specifically, Figures 21A and 21B show time-domain TBEV and ACC waveforms simultaneously measured from a patient using a system similar to that described above. In the TBEV waveform, the slowly varying pulses that occur approximately every seven seconds correspond to individual respirations, while the sharp peaks in the waveform correspond to pulses induced by the heartbeat. Figures 21C and 21D show the frequency-domain power spectra of both the TBEV and ACC waveforms, respectively. A major peak near 0.8 Hz is clearly visible in the power spectrum of the TBEV waveform, corresponding to the heartbeat-induced pulse. A much weaker peak corresponding to the patient's respiratory rate is also evident near 0.15 Hz. As shown in the gray-shaded area 99, the power spectrum corresponding to the ACC waveform is characterized by a single, well-defined peak near 1.5 Hz, which includes frequency components that are nearly identical to the corresponding peak in the TBEV waveform. Further processing of these two spectra using a simple peak-finding algorithm yields the patient's actual respiratory rate, which corresponds to approximately 8 breaths per minute.

Measuring TBEV waveforms in the absence of motion

Figures 22A, 22B and 23A, 23B illustrate how varying degrees of motion from a patient's arm can affect both the ECG and TBEV waveforms, thereby affecting the accuracy of SV measurements. The ACC waveform is typically measured along the vertical axis of an accelerometer embedded in the TBEV module. The amplitudes of the axes of the ECG and TBEV waveforms are the same for all figures.

In FIG22B , for example, the ACC waveform is relatively flat and lacks any significant time-dependent features, indicating that the patient is not moving and is relatively still. Thus, the TBEV waveform in FIG22A , which is strongly affected by motion, is characterized by a well-defined value for pulse onset, indicated by marker 73, and _{a maximum value for} (dZ/dt), indicated by marker 74. Similarly, the ECG waveform is characterized by an undistorted QRS complex, indicated by marker 72. The fidelity of these features indicates that both HR and SV values can generally be accurately determined during periods of little or no motion, as indicated by the ACC waveform in FIG22B .

Figures 23A and 23B illustrate the effects of extensive arm movement on both the ECG and TBEV waveforms. Here, periods of movement are indicated by dashed boxes 80 in both figures, contrasting with preceding periods in which there was no movement, as shown in dashed boxes 81. The ACC waveform in Figure 23B shows that the movement lasted approximately one second, beginning and ending near the times indicated by markers 78 and 79, respectively. The movement was complex, and its intensity peaked at marker 79. Even for major finger movements, the ECG waveform and its QRS complex (indicated by marker 75) were relatively undistorted. However, the TBEV measured during the movement period was strongly distorted, so that its peak (indicated by marker 77) was relatively flat and essentially unmeasurable. This makes it difficult to accurately measure the TBEV waveform and subsequently calculate the SV value from this parameter. The peak onset (indicated by marker 76) was also distorted, but not to the same extent as the corresponding peak.

The data shown in Figures 22A and 22B demonstrate that motion can be detected and accounted for during TBEV measurements to minimize false alarms and, in addition, provide accurate readings in the presence of motion. For example, during periods of motion, the SFT can be calculated using Weissler regression and then used in Equation 3 above to estimate the SV. Alternatively, during such periods of motion, the SV can be estimated using one of the various estimators shown in Figure 1.

Processing ACC waveforms to determine posture

A patient's posture can affect their SV/CO/CP values, so understanding this parameter can improve the measurements described herein. To perform this measurement, the body-worn monitor includes three 3-axis accelerometers, as well as ECG and TBEV circuitry. In addition to measuring SV/CO/CP, these sensors generate time-dependent waveforms that, when analyzed, yield RR and the patient's motion-related properties, such as exertion, posture, and activity level.

Figures 25A-25B illustrate, for example, that a body-worn monitor can generate an ACC waveform that can be analyzed to accurately estimate a patient's posture. Specifically, Figure 25A shows that a 3-axis accelerometer in the torso accurately measures the ACC waveform, which directly correlates to the patient's position (e.g., standing, supine, prone, or side-lying) (as shown in Figure 25B). This capability, in turn, can be used along with the patient's SV/CO/CP values and a series of "heuristics" to generate alarm/alert values. For example, if a patient's SV/CO/CP values are low (e.g., below approximately 1.5 liters/min for CO), but analysis of the ACC waveform indicates that the patient is standing or walking, as shown in Figure 25B, then no alarm is required. The assumption here is that a patient in this posture/activity level does not require medical assistance. Conversely, the combination of low SV/CO/CP values and a patient lying supine, or worse, recently fallen, should trigger an alarm.

FIG24 illustrates how a body-worn monitor can determine motion-related parameters (e.g., degree of motion, posture, and level of motion) from a patient 410 using time-dependent ACC waveforms continuously generated by three accelerometers 412, 413, and 414 worn on the patient's chest, bicep, and wrist, respectively. Furthermore, the height of the patient's arm may also affect cNIBP measurements, as blood pressure can vary significantly due to the hydrostatic pressure induced by changes in arm height. Furthermore, this phenomenon can be detected and utilized to calibrate cNIBP measurements, as described in detail in the aforementioned patent applications, the contents of which have been previously incorporated by reference. As described in these documents, arm height can be determined using DC signals from accelerometers 413 and 414, respectively, placed on the patient's bicep and wrist. In contrast, posture can be determined exclusively by accelerometer 412 worn on the patient's chest. An algorithm operating on the wrist-worn transceiver extracts DC values from the waveforms measured by this accelerometer and processes them using the algorithm described below to determine posture.

Specifically, for patient 410, torso posture is determined using the angles measured between the measured gravity vector and the axes of torso coordinate space 411. The axes of this space 411 are defined in three-dimensional Euclidean space, where is the vertical axis, is the horizontal axis, and is the normal axis. These axes must be identified relative to the "chest accelerometer coordinate space" before the patient's posture can be determined.

The first step in determining the patient's posture is identifying its alignment in the chest accelerometer coordinate space. This can be determined in one of two ways. In the first approach, an assumption is made based on the typical alignment of the body-worn monitor relative to the patient. During the manufacturing process, these parameters are then pre-programmed into the firmware running on the wrist-worn transceiver. This program assumes that the accelerometers within the body-worn monitor are used in a substantially identical configuration for each patient. In the second approach, this is identified on a patient-specific basis. Here, an algorithm operating on the wrist-worn transceiver prompts the patient (using, for example, video instructions running on the wrist-worn transceiver or audio instructions transmitted through a speaker) to assume a known position relative to gravity (e.g., standing upright with the arm pointing straight down). The algorithm then calculates the patient's position from the DC values corresponding to the chest accelerometer's x-, y-, and z-axes when the patient is in this position. However, this still requires knowing which arm (left or right) the monitor is worn on, as the chest accelerometer coordinate space can be rotated 180 degrees depending on this orientation. The medical professional applying the monitor can enter this information using the aforementioned GUI. This dual-arm attachment option requires a set of two predetermined perpendicular and normal vectors that are interchangeable depending on the position of the monitor. Instead of manually entering this information, the arm on which the monitor is worn can be easily determined after attachment using the values measured by the chest accelerometer, assuming non-orthogonality to the gravity vector.

The second step of the procedure is to identify the alignment in the chest accelerometer coordinate space. The monitor determines this vector in the same way it determines it using one of two methods. In the first method, the monitor assumes a typical alignment of the chest-worn accelerometer on the patient. In the second method, the patient is prompted to perform the alignment procedure and is asked to assume a known position relative to gravity. The monitor then calculates the DC value of the time-dependent ACC waveform.

The third step of the procedure is to identify the alignment of in the chest accelerometer coordinate space.This vector is typically determined by the vector cross product of and , or it can be assumed based on the typical alignment of the accelerometers on the patient, as described above.

The patient's posture is determined using the above coordinate system and (in Figure 25) a gravity vector extending normal from the patient's chest. The angle between and is given by Equation 7:

where the dot product of these two vectors is defined as:

The definition of the norm of the sum is given by Equations 9 and 10:

As shown in Equation 12, the monitor compares the vertical angle θ _VG to a threshold angle to determine whether the patient is vertical (i.e., standing upright) or lying down:

If θ _VG ≤ 45°, then trunk state = 0, the patient is upright (11)

If the conditions in Equation 11 are met, the patient is assumed to be upright and their trunk state (which is a numerical value equivalent to the patient's posture) is equal to 0. If θ _VG > 45 degrees, the patient is assumed to be lying down. Their recumbent position is then determined by the angle between the two remaining vectors, as defined below.

The angle θ _NG between and determines whether the patient is in the dorsal (chest up), prone (chest down), or lateral position. Based on the assumed orientation or the patient-specific calibration procedure (described above), the alignment is given by Equation 11, where i, j, k denote the unit vectors of the x-, y-, and z-axes of the chest accelerometer coordinate space, respectively:

The angle between and measured from the DC value extracted from the chest ACC waveform is given by Equation 13:

The body-worn monitor measures the normal angle θ _NG and then compares it to a set of predetermined threshold angles to determine which lying position the patient is in, as shown in Equation 14:

If the conditions in Equation 14 are not met, then the patient is assumed to be side-lying. Whether they are right or left-side decubitus is determined by the angle calculated between the horizontal torso vector and the measured gravity vector as described above.

The alignment of is determined using an assumed orientation, or by the vector cross product of and given by Equation 15, where i, j, and k represent the unit vectors of the x-, y-, and z-axes of the accelerometer coordinate space, respectively. Note that the orientation of the calculated vector depends on the order of the vectors in the calculations. The following order defines the horizontal axis as being toward the right side of the patient's body.

The angle θ _HG between and is determined using Equation 16:

The monitor compares this angle to a set of predetermined threshold angles to determine whether the patient is lying on the right or left side, as given by Equation 17:

Table 1 describes each of the above gestures, along with a corresponding numerical torso state used to provide, for example, a specific icon on a remote computer:

Stand upright 0 Supine: back facing down 1 Prone: chest down 2 Lying on the right side 3 Lying on the left side 4 Undetermined posture 5

Table 2 — Postures and their corresponding trunk states

The data shown in Figures 25A and 25B were calculated using the method described above. As the patient moves, the DC value of the ACC waveform measured by the chest accelerometer changes accordingly, as shown in Figure 25A . The body-worn monitor processes these values as described above to continuously measure various quantized torso states for the patient, as shown in Figure 25B . The torso states yield the patient's posture as defined in Table 2. For this study, the patient rapidly alternated between standing, lying on their back, lying on their stomach, lying on their right side, and lying on their left side over a period of approximately 160 seconds. Different alarm/alert conditions (e.g., thresholds) for vital signs can be assigned to each of these postures, or specific postures themselves can cause an alarm/alert. Additionally, the time-dependent performance of the graphs can be analyzed (e.g., by counting changes in torso state) to determine, for example, how often a patient moves in their bed. This number can then be considered equivalent to various metrics, such as a "decubitus ulcer index," which indicates how immobile a patient is in their bed, potentially causing lesions.

Alternative implementation plans

Other embodiments are within the scope of the present invention. For example, the TBEV harness and its associated electrodes can assume a variety of configurations. One of these configurations is illustrated in FIG27 . Here, the TBEV harness 150 features a TBEV module 156, which is positioned directly on top of a single TBEV electrode 158. Electrode 158 features four conductive regions (not shown): 1) a current source; 2) a current sink; and 3) and 4) a pair of electrodes for measuring voltage in the CB region. As described above, the conductive regions for sourcing and sinking current are located on the outer portion of electrode 158, while those for measuring voltage are located on the inner portion. Each conductive region is connected to analog circuitry within TBEV module 156 via a single connector (not shown). TBEV module 156 also includes a CAN transceiver (not shown) that transmits digitized waveforms and CO/SV values via a first cable 154 to connector 152, which plugs into the back panel of the wrist-worn transceiver, as shown in FIG10 . A second cable 160 connects to an ECG module 162, which in turn connects to a collection of ECG leads 166 via a short third cable 164. During a measurement, the ECG module 163 transmits a digitized version of the ECG waveform, HR, and other information via the second cable 160 and to the TBEV module 156. Data is transmitted according to the CAN protocol. From there, the data is relayed by the module's internal CAN transceiver through the second cable 154 and to the connector 152, which then passes the data to the wrist transceiver.

FIG28 shows an alternative embodiment of the present invention in which the TBEV 449 and ECG 420 modules are physically separated and connected via a wireless interface. Here, the ECG module 420 includes ECG circuitry and is attached to ECG electrodes 424a to 424c via cables 430a to 430c. A second arm-worn module 449 includes four electrodes (two for injecting current and two for measuring voltage) dispersed across its upper and lower portions, which are connected to the central TBEV circuitry to perform measurements on the brachial plane as described above. Both the chest-worn module 420 and the arm-worn module 449 include unique Bluetooth transmitters that send ECG and TBEV waveforms, respectively, to paired Bluetooth transmitters 128 in the wrist-worn transceiver 426.

FIG29 illustrates the arm-worn module 449 in greater detail. As described above, it includes four electrodes 448a-448d, which are attached to the back of a flexible substrate 451 holding the TBEV circuitry 447, located in the center of the module. Electrodes 448a-448d, as described above, provide the current injection and voltage measurement functions of the TBEV measurement and are connected to the TBEV circuitry 447 via a series of metal traces 453a-453d embedded within the flexible substrate. Electrodes 448a-448d also adhere to the patient's skin to secure the module 449 to the arm. Once the TBEV waveform is measured, the Bluetooth transmitter 446 located at the bottom of the module sends it to the wrist-worn transceiver for processing, as described above.

Figures 30 and 31 show schematic diagrams of an alternative embodiment of the present invention and illustrate how data related to SV/CO/CP can be wirelessly transmitted from the chest-worn sensor 500 and wrist-worn transceiver 506 to an external router (517a in Figure 30 and 517b in Figure 31), and from there to an external network (e.g., the Internet). This data transmission process can use a variety of strategies, two of which are shown in the figure. In Figure 30, for example, ICG and ECG waveforms are measured from analog signals collected by electrode patches 502a, 502b, 404, and then wirelessly transmitted from the chest-worn sensor 500 to the wrist-worn transceiver 506 using Bluetooth, where they are then analyzed as described above to determine SV/CO/CP, as well as all other vital signs. This processed data is then sent from the transceiver 506 to the external router 517a using Bluetooth, 802.11, or any other wireless protocol. Once the data is received by router 506, it transmits the data to an external network using a wireless protocol (e.g., CDMA, GSM, iDEN) or a wired protocol (e.g., Ethernet). From there, the data can be transmitted to a hospital medical record system, a website, or sent to another application via a web service.

FIG31 illustrates an alternative approach in which an external router 517b performs a higher degree of computational load. In this scenario, the chest-worn sensor 500 processes the analog signals measured by the electrode patches 502a, 502b, and 504 to determine ECG and ICG waveforms, then wirelessly transmits these analog signals in digital form to the router 517b. Nearly simultaneously, the wrist-worn transceiver measures SpO2 and PPG waveforms and wirelessly transmits these waveforms to the router 517b. There, an embedded processor analyzes the ECG waveform to determine HR; analyzes the ECG and PPG waveforms to determine PAT and cNIBP; and analyzes the ECG, ICG, PPG waveforms, and PAT to determine SV/CO/CP. This data is then transmitted to an external network as described above and from there to another system.

In addition to those described above, body-worn monitors can use many other methods to calculate blood pressure and other properties from optical and electrical waveforms. These methods are described in the following co-pending patent applications, the contents of which are incorporated herein by reference: 1) CUFFLESS BLOOD-PRESSURE MONITOR AND ACCOMPANYING WIRELESS, INTERNET-BASED SYSTEM (U.S.S.N. 10/709,015; filed April 7, 2004); 2) CUFFLESS SYSTEM FOR MEASURING BLOOD PRESSURE (U.S.S.N. 10/709,014; filed April 7, 2004); 3) CUFFLESS BLOOD PRESSURE MONITOR AND ACCOMPANYING WEB SERVICES INTERFACE (U.S.S.N. 10/810,237; filed March 26, 2004); 4) VITAL SIGN MONITOR FOR ATHLETIC APPLICATIONS (U.S.S.N; filed September 13, 2004); 5) CUFFLESS BLOOD PRESSURE MONITOR AND ACCOMPANYING WIRELESS MOBILE DEVICE (U.S.S.N. 10/967,511; filed October 18, 2004); 6) BLOOD PRESSURE MONITORING DEVICE FEATURING A CALIBRATION-BASED ANALYSIS (U.S.S.N. 10/967,610; filed October 18, 2004); 7) PERSONAL COMPUTER-BASED VITAL SIGN MONITOR (U.S.S.N. 10/906,342; filed February 15, 2005); 8) PATCH SENSOR FORMEASURING BLOOD PRESSURE WITHOUT A CUFF (U.S.S.N. 10/906,315; filed February 14, 2005); 9) PATCH SENSOR FOR MEASURING VITAL SIGNS (U.S.S.N. 11/160,957; filed July 18, 2005); 10) WIRELESS, INTERNET-BASED SYSTEM FOR MEASURING VITAL SIGNS FROM APLURALITY OF PATIENTS IN A HOSPITAL OR MEDICAL CLINIC (U.S.S.N. 11/162,719; filed September 9, 2005); 11) HAND-HELD MONITOR FOR MEASURING VITAL SIGNS (U.S.S.N. 11/162,742; filed September 21, 2005); 12) CHEST STRAP FOR MEASURING VITAL SIGNS (U.S.S.N. 11/306,243; filed December 20, 2005); 13) SYSTEM FOR MEASURING VITAL SIGNS USING AN OPTICAL MODULE FEATURING A GREEN LIGHT SOURCE (U.S.S.N. 11/307,375; filed February 3, 2006); 14) BILATERAL DEVICE, SYSTEM AND METHOD FOR MONITORING VITALSIGNS (U.S.S.N. 11/420,281; filed May 25, 2006); 15) SYSTEM FOR MEASURING VITALSIGNS USING BILATERAL PULSE TRANSIT TIME (U.S.S.N. 11/420,652; filed May 26, 2006); 16) BLOOD PRESSURE MONITOR (U.S.S.N. 11/530,076; filed September 8, 2006); 17) TWO-PART PATCH SENSOR FOR MONITORING VITAL SIGNS (U.S.S.N. 11/558,538; filed November 10, 2006); and 18) MONITOR FOR MEASURING VITAL SIGNS AND RENDERING VIDEO IMAGES (U.S.S.N. 11/682,177; filed March 5, 2007).

Other embodiments are also within the scope of the present invention. For example, other measurement techniques (such as conventional oscillometric measurement during deflation) can be used to determine SYS for the above-described algorithm. In addition, the processing unit and probe (similar to those described above) for measuring SpO2 can be modified and worn on other parts of the patient's body. For example, an optical sensor with a ring configuration can be worn on a finger other than the thumb. Or they can be modified to attach to other conventional sites for measuring SpO2, such as the ear, forehead, and bridge of the nose. In these embodiments, the processing unit can be worn somewhere other than the wrist, such as around the neck (and (for example) supported by a lanyard) or on the patient's waist (for example, supported by a clip attached to the patient's belt). In yet other embodiments, the probe and processing unit are integrated into a single unit.

In other embodiments, a group of body-worn monitors can continuously monitor a group of patients, wherein each patient in the group wears a body-worn monitor similar to those described herein. In addition, each body-worn monitor can be supplemented with a position sensor. The position sensor includes a wireless component and a position processing component that receives a signal from the wireless component and processes it to determine the patient's physical location. The processing component (similar to the processing component described above) determines at least one vital sign, a motion parameter, and an alarm parameter calculated from the combination of this information from the time-dependent waveform. A wireless transceiver transmits the patient's vital signs, motion parameter, position, and alarm parameter via a wireless system. A remote computer system (featuring a display and interface to the wireless system) receives the information and displays the information on a user interface for each patient in the group.

In an embodiment, the interface provided on the display at the central nursing station is characterized by a map field that displays a map corresponding to an area having a plurality of segments. Each segment corresponds to the location of a patient and includes, for example, the patient's vital signs, motion parameters, and alarm parameters. For example, the map field may display a map corresponding to an area of a hospital (e.g., a medical office or emergency room), wherein each segment corresponds to a specific bed, chair, or approximate location in the area. Typically, the display provides graphical icons that correspond to the motion and alarm parameters of each patient in the group. In other embodiments, the body-worn monitor includes a graphical display that provides these parameters directly to the patient.

Typically, the location sensor and wireless transceiver operate on a common wireless system (e.g., one based on 802.11 (i.e., "WiFi"), 802.15.4 (i.e., "Bluetooth"), or cellular (e.g., CDMA, GSM) protocols). In this case, location is determined by processing wireless signals using one or more algorithms known in the art. These algorithms include, for example, triangulating signals received from at least three different base stations, or estimating location based solely on signal strength and proximity to a particular base station. In yet other embodiments, the location sensor includes a conventional Global Positioning System (GPS) that processes signals from orbiting satellites to determine the patient's location.

The body-worn monitor may include a first voice interface, while the remote computer may include a second voice interface that is integrated with the first voice interface. The position sensor, wireless transceiver, and first and second voice interfaces may all operate on a common wireless system (such as one of the systems described above) based on 802.11 or cellular protocols. The remote computer may, for example, be a monitor substantially identical to the monitor worn by the patient and may be carried or worn by the medical professional. In this case, the monitor associated with the medical professional may feature a GUI in which the user can select to display information corresponding to a particular patient (e.g., vital signs, location, and alarms). This monitor may also include a voice interface, allowing the medical professional to communicate directly with the patient.

Still further implementations are within the scope of the following claims.

Claims (6)

1. A system for measuring stroke volume of a patient, comprising: an impedance sensor comprising at least two electrodes and an impedance circuit, the at least two electrodes being configured to be positioned in the clavicle/brachial region of the patient, and the impedance circuit being configured to receive and process signals from the at least two electrodes to measure a transbrachial electrical velocimetry (TBEV) signal from the patient; an optical sensor connected to an optical probe attached to the patient and comprising optical circuitry configured to receive and process signals from the optical probe to measure at least one optical signal from the patient; and a processing system configured to: i) be worn on the patient's body; ii) be operably connected to both the impedance sensor and the optical sensor; iii) receive the TBEV signal from the impedance sensor; iv) receive the at least one optical signal from the optical sensor; v) process the TBEV signal to determine a time-dependent derivative dZ/dt of the TBEV signal, a reference point B in dZ/dt representing an opening of the patient's aortic valve, a reference point X in dZ/dt representing a closing of the patient's aortic valve, and a value of (dZ/dt) _maximum , wherein (dZ/dt) _maximum represents ΔZ(t ) the maximum slope when the pulse rebounds; (vi) calculating the first systolic flow time SFT value by the time differential between the reference point B and the reference point X; (vii) centrally processing the brachial base impedance Zo, _{the maximum value of} (dZ/dt) and the first SFT value to determine a first stroke volume per stroke SV value; (viii) processing the optical signal to determine a second SFT value; (ix) centrally processing Zo, _{the maximum value} of (dZ/dt) and the second SFT value to determine a second SV value; (x) centrally processing the first SV value and the second SV value to determine a third SV value; (xi) displaying the third SV value on a display device. 2. The system for measuring a patient's stroke volume as recited in claim 1 , wherein the impedance sensor comprises a housing configured to be worn on the patient's body. 3. The system for measuring the stroke volume of a patient of claim 1 , wherein the impedance sensor comprises an analog-to-digital converter configured to convert the TBEV signal into a digital TBEV signal. 4. The system for measuring a patient's stroke volume as recited in claim 3, wherein the impedance sensor comprises a transceiver configured to transmit the digital TBEV signal to the processing system via a cable. 5. The system for measuring stroke volume of a patient as recited in claim 3, wherein the impedance sensor comprises a transceiver configured to transmit the digital TBEV signal to the processing system via a wireless interface. 6. A system for measuring a patient's stroke volume as described in claim 5, wherein the transceiver comprises a short-range radio transceiver.