Systems For Synchronizing Different Devices To A Cardiac Cycle And For Generating Pulse Waveforms From Synchronized ECG and PPG Systems

Abstract

A system for synchronizing a target device to a cardiac cycle, including: (a) a target device that collects data or performs an operation that is to be timed to the cardiac cycle; (b) a signaling device that emits a signal indicating the occurrence of a cardiac contraction and/or ECG feature; and (c) a calibration device that determines the relationship of the signal from the signaling device to the actual cardiac cycle. In operation, the calibration device calculates a time offset between the timing of the cardiac contraction as determined by the signaling device and the timing of the cardiac contraction and/or ECG feature as determined by the calibration device, and then provides the time offset to the target device.

Claims

1. A system for synchronizing a first target device to a cardiac cycle, comprising: (a) a first target device that performs an operation that is timed to a cardiac cycle; (b) a signaling device that emits a signal indicating the occurrence of a cardiac contraction; and (c) a calibration device that determines the timing of the cardiac cycle, wherein the calibration device receives the signal from the signaling device and calculates a time offset between the timing of the signal from the signaling device and the timing of the cardiac cycle as determined by the calibration device, and wherein the calibration device provides the time offset to the first target device thereby enabling synchronization of the first target device to the cardiac cycle.

2. The system of claim 1, wherein the time offset provided to the first target device comprises an adjustment to an internal clock in the target device.

3. The system of claim 1, wherein the signaling device emits a signal having a fixed consistent time relationship to an actual heart contraction.

4. The system of claim 3, wherein the signal emitted by the signaling device identifies points in time in the cardiac cycle that are not times of heart contraction.

5. The system of claim 1, wherein the first target device is one of: a PPG system; a cardiac/blood property monitoring device; a drug delivery device; a fluid sampling device; a fluid measuring device; a robotic surgery device; an imaging device; or a pacemaker.

6. The system of claim 1, wherein the signaling device is: a heart rate measuring device, an ECG system, an imaging device, including but not limited to a fluoroscope, video-camera, MRI or CT machine, an acoustic device, including but not limited to a stethoscope, or a physical sensing device capable of determining a heart contraction, including but not limited to a chest belt strap device.

7. The system of claim 1, further comprising: (d) a second target device that performs an operation that is timed to the cardiac cycle, wherein the calibration device provides the time offset to the second target device thereby enabling synchronization of the second target device to the cardiac cycle.

8. The system of claim 7, wherein the first and second target devices are both PPG systems configured to be positioned on different anatomical locations on a patient.

9. The system of claim 1, wherein: the first target device comprises a PPG system, the signaling device comprises an ECG system, and the calibration device comprises an ECG system, and wherein the calibration device is in communication with the first target device, and the signaling device is in communication with the first target device.

10. The system of claim 9, wherein the signaling ECG system shares the internal clock of the calibration ECG system.

11. The system of claim 9, wherein the time offset is the difference in time of the detection of a QRS signal between each of the calibration and signaling ECG systems.

12. The system of claim 1, wherein the leads of the signaling or calibration device are disposed in opposite sides or ends of a hand-held device or hand-held device cover.

13. The system of claim 1, wherein the leads of the calibration device are disposed in a patch worn on a person's skin.

14. The system of claim 1, wherein the calibration device is removed after the time offset has been provided to the first target device.

15. The system of claim 1, wherein the signal emitted by the first target device is a composite PPG signal.

16. The system of claim 15, wherein the composite PPG signal is generated by comparing segments of a PPG signal taken over a plurality of cardiac cycles.

17. The system of claim 16, wherein comparing the segments of the PPG signal comprises comparing segment lengths of that segment or prior segment to one another and then sorting segments of similar length into categories and then generating composite signal segments for each of the categories.

18. A system for synchronizing a first target device to a cardiac cycle, comprising: (a) a first target device that performs an operation that is timed to a cardiac cycle; and (b) a calibration-and-signaling device that determines the timing of the cardiac cycle, wherein the calibration-and-signaling device calculates a time offset between the timing of the occurrence of the cardiac contraction and the timing of the cardiac cycle as determined by the calibration-and-signaling device, and wherein the calibration-and-signaling device provides the time offset to the first target device thereby enabling synchronization of the first target device to the cardiac cycle.

19. The system of claim 18, wherein the signal emitted by the first target device is a composite PPG signal.

20. The system of claim 19, wherein the composite PPG signal is generated by comparing segments of a PPG signal taken over a plurality of cardiac cycles.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0022] FIG. 1 is a schematic illustration of the present system.

[0023] FIG. 2 is an illustration of the signal readings of the various components of a preferred embodiment of the present system.

[0024] FIG. 3A is an illustration of a preferred embodiment of the present system using a removable calibration device (prior to removal of the calibration device).

[0025] FIG. 3B is an illustration corresponding to FIG. 3B, but with the calibration device removed.

[0026] FIG. 4A is an illustration of an exemplary calibration ECG system positioned on a patient's chest, with an exemplary signaling ECG system and target PPG system disposed in a band around the patient's arm.

[0027] FIG. 4B is a sectional view through the patient corresponding to FIG. 4A.

[0028] FIG. 5A is a top perspective view of an exemplary signaling and target device disposed in an adhesive chest patch.

[0029] FIG. 5B is a bottom perspective view of the exemplary signaling and targeting device of FIG. 5A.

[0030] FIG. 5C illustrates exemplary signaling and targeting devices disposed in a chest strap worn by the patient.

[0031] FIG. 6A is an exemplary handheld signaling and target device positioned on a patient's chest.

[0032] FIG. 6B is a sectional view through the patient corresponding to FIG. 6A.

[0033] FIG. 7A is a top perspective view of an exemplary signaling and targeting device that is held in a patient's hands.

[0034] FIG. 7B is a bottom perspective view of the signaling device of FIG. 7A.

[0035] FIG. 8A illustrates ECG and PPG signals measured over a plurality of cardiac cycles.

[0036] FIG. 8B illustrates the generation of a PPG composite signal from averaged PPG signal segments corresponding to successive cardiac cycles.

[0037] FIG. 9 is a first side-by-side comparison of ECG and PPG signals showing steps in an alternate method of generating a PPG composite wave.

[0038] FIG. 10 is a second side-by-side comparison of ECG and PPG signals showing steps in an alternate method of generating a PPG composite wave.

DETAILED DESCRIPTION OF THE DRAWINGS

[0039] FIG. 1 is a schematic illustration of the present system 10 for synchronizing one or more target devices T1, T2 . . . Tn to a cardiac cycle. System 10 comprises: (a) at least one target device T1 (T2 . . . to Tn) that performs an operation that is to be timed to a cardiac cycle; (b) a signaling device S that emits a signal indicating the occurrence of a cardiac contraction; and (c) a calibration device C that determines the timing of the cardiac cycle. As will be explained herein with reference to FIG. 2, the calibration device C receives a signal from signaling device S and then calibration device C calculates a time offset TO between the timing of the heart contraction as determined by the signaling device S and the timing of the heart contraction in the cardiac cycle as determined by the calibration device C. Next, as will also be fully explained, the calibration device C provides the time offset TO to target device T1 thereby enabling synchronization of target device T1 to the cardiac cycle.

[0040] In optional preferred embodiments of the present system, the first target device T1 (and various additional target devices T2 to Tn) may each be one of the following systems or devices: a PPG (photoplethysmography) system; any cardiac/blood property monitoring device; a drug delivery device; a fluid sampling device; a fluid measuring device; a robotic surgery device; an imaging device; or a pacemaker. It is to be understood, however, that the present target device T1 to Tn are not limited to only to these specific devices. It is also to be understood that the present system encompasses embodiments with only one target device T1, and embodiments with any plurality of target devices T1 to Tn.

[0041] In optional preferred embodiments of the present system, the signaling device S may be one of the following systems and devices: a heart rate measuring device, an ECG system, an imaging device, including but not limited to a fluoroscope, video-camera, MRI or CT machine, an acoustic device, including but not limited to a stethoscope, or a physical sensing device capable of determining a heart contraction, including but not limited to a chest belt strap device.

[0042] Optionally, a plurality of target devices T1 to Tn each perform an operation that is to be accurately timed to the cardiac cycle, and the calibration device C provides the time offset TO to each of these target devices, thereby enabling synchronization of each of the plurality of target devices to the cardiac cycle.

[0043] In one preferred embodiment of the present system (further explained in FIG. 2 and FIGS. 3A and 3B), the signaling device S is a first ECG system, the calibration device is a second ECG system and the target device is a PPG system. (Preferably, the signaling device S can be a simple, inexpensive ECG system such as a system in a wrist watch or band, or in a chest band; whereas the calibration ECG system C can be a more expensive and more accurate ECG system).

[0044] In another preferred embodiment of the present system, the signaling device S is an MRI system, the calibration device is an ECG system and the target device is a Doppler system. Optionally, the signaling device S may instead be an MRI in Cine mode or an Echocardiogram system, and the target device may instead be a PPG system.

[0045] In each of these various embodiments above, the calibration device C is in communication with the target device T1, and the signaling device S is also in communication with target device T1.

[0046] In optional embodiments of the present system, the calibration and signaling devices are combined into an integrated device that performs both functions. As such, a system for synchronizing a first target device to a cardiac cycle is provided, comprising: (a) a target device(s) that performs an operation that is timed to a cardiac cycle; and (b) a calibration-and-signaling device that determines the timing of the cardiac cycle. In this embodiment, the calibration-and-signaling device calculates a time offset between the timing of the occurrence of the cardiac contraction and the timing of the cardiac cycle as determined by the calibration-and-signaling device, and the calibration-and-signaling device provides the time offset to the first target device thereby enabling synchronization of the first target device to the cardiac cycle.

[0047] FIG. 2 is an illustration of the signal readings of the various components of one preferred embodiment of the present system, as provided by the various components of an exemplary embodiment of the present system (as further illustrated in FIGS. 3A and 3B), as follows.

[0048] The signal emitted by signaling device S is a repeating waveform generally corresponding to the user's cardiac cycle, showing the times at which the heart's QRS wave is detected. Similarly, the signals detected by calibration device C is also a repeating waveform generally corresponding to the user's cardiac cycle, also showing the times at which the heart's QRS wave is detected. As can be seen, the signaling and calibration devices S and C do not detect the heart's QRS wave at exactly the same times. This is due to the fact that the signaling device S may be a cheaper, simpler device having an inherent signal time delay (as compared to the more sophisticated calibration device C). In addition, the delay in the signal from signaling device S results both from the combination of the delay in the circuit itself (i.e.: the time spent for signaling device S to read and transmit its signal) and the delay in the signal traveling across the body (for example, the signal traveling from a different body location from that of calibration system C).

[0049] As can be seen, the time offset TO is the difference in time of the detection of a QRS signal between each of the calibration and signaling systems C and S. In accordance with the present invention, the calibration system C senses the cardiac cycle, and knowing its own delay properties it determines the time offset TO that is then provided to target device T1 so that the target device T1 can synchronize to the cardiac cycle.

[0050] In one preferred method, the time offset TO provided to the first target device T1 comprises an adjustment to be made to the internal clock output in target device T1. As such, the time offset TO provided to the first target device T1 may either comprise an adjustment of the times reported by the first target device when sensing specific physiological features of the cardiac cycle, or the times of performing actions based on specific physiological features of the cardiac cycle.

[0051] In the preferred exemplary aspect illustrated in FIGS. 3A and 3B, the time offset TO provided by calibration system C will be used to permit target device T1 to accurately measure a patient or user's PWTT (so as to generate various pulse metrics). It is to be understood, however, that many other applications of the present system are also contemplated within the scope of the present invention.

[0052] In various aspects of the present system, the signaling device S emits a signal that either: has a fixed consistent time relationship to an actual heart contraction, or is not specific as to cardiac cycle phase. For example, the signaling device may emit a simple “beep” only at points in time when it senses a heart contraction, or it may emit a continuous signal that corresponds to other known points in a cardiac cycle that are not times of heart contraction. In the illustration of FIG. 2, the signaling device emits a continuous ECG signal.

[0053] Typically, for the signaling device S, it is expected that the signal processing delays (i.e.: delays within the circuitry itself) will be the major delay factor and that delays caused by individual patient physiology (i.e.: the speed of travel of electrical signals through the patient's body) will be small. The speed at which signals travel through the patient's body can vary over time as the patient's health changes. Also, different types of signaling devices S will have different delays. All of these delays will be consistent for one patient with one set of devices at one time. The present system can effectively deal with all these irregularities since it relies upon a more accurate calibration ECG system C to determine the exact timing of the cardiac cycle.

[0054] In the embodiments illustrated in FIGS. 3A and 3B, the first target device T1 comprises a first PPG system, the second target device T2 comprises a second PPG system, the signaling device S comprises a first (simple, less accurate) ECG system, and the calibration C device comprises a second (more accurate) ECG system. The calibration device C is in communication with the target devices T1 and T2, and the signaling device S is also in communication with the target devices T1 and T2.

[0055] In the illustrated embodiment, the first and second target devices T1, T2 are both PPG systems configured to be positioned on different anatomical locations on a patient, for example, the opposite lateral limbs of a patient (e.g.: fingers on the patient's left and right hands).

[0056] The objective of the system illustrated in FIGS. 3A and 3B is to easily calculate the patient's simultaneous PWTT to each of the patient's opposite limbs. (It is to be understood that target device T2 can be removed from FIGS. 3A and 3B so that the system instead functions only to calculate the PWTT to one limb at a time).

[0057] A user can keep track of their personal fitness by monitoring the pulse metrics obtainable once a stable/reproducible PWTT is established for any given scenario. Such pulse metrics (shape/slope/peaks/rolloff, etc.) provide insight into the cardiovascular status of the individual, such as whether peripheral arterial resistance is high or low.

[0058] In the illustrated embodiment of FIGS. 2, 3A and 3B, PWTT is determined by measuring the time difference between the onset of the heart contraction (i.e.: the accurate time detection of the QRS signal as measured by the calibration ECG system C) and the time at which the peak arterial pulse reaches a desired location on the patient's body (i.e.: the accurate time detection of the maximum and minimum of the signal reading taken by a PPG device T1 at a patient's finger tips). A PPG (photoplethysmography) system measures changes in the light reflected from or transmitted through the illuminated skin. The blood pulse wave distends the arterioles as it passes through them. Therefore, the arrival time of each pulse in the cardiac cycle can be read as a maximum (the onset of the arterial pulse) and a minimum (at the peak of the pulse) in the signal from the PPG's light sensor.

[0059] A major problem with using existing ECG and PPG systems together is that they typically each have their own dedicated internal clocks which measure time separately. As such, synchronizing ECG and PPG time signals has proven to be especially problematic because of the effect of very small (microsecond to millisecond) differences in clock timing. These problems occur even with signal time differences even being a few microseconds or milliseconds apart. In addition, problems also occur with simple ECG signaling systems due to the high noise to signal ratio and potential for outside interference. Measuring a patient's ECG with a simple fitness tracker signaling device is also problematic due to intermittent connections inherent in poor skin connection. Motion of the patient also degrades the accuracy when taking an ECG reading with a simple device. Moreover, the most accurate ECG readings are taken when the ECG leads are positioned far apart on the patient. As such, the most accurate ECG measurement approaches tend to be the ones that are most intrusive, or require the patient to remain motionless in a hospital or doctor's office. It would instead be desirable to provide an accurate, synchronized ECG system that can be used while moving or exercising. The present solution addresses these concerns and enables a person to simply, cheaply (and accurately) measure their own arterial pulse metrics in the convenience of their own home or place of exercise.

[0060] Prior art solutions instead often relied on a (3.sup.rd) master clock to send time signals to each of the internal clocks of the ECG and PPG monitoring systems. Objectives of the present system are to achieve time synchronization: (a) without relying on a 3rd master clock, (b) without relying on a 2nd separate clock timing in one of the ECG or PPG systems, and (c) without having to determine which of two clocks is “more correct”, and then make adjustments or apply some form of averages to these multiple clocks.

[0061] Another objective of the present system is the removal of the wired connection between the ECG and PPG monitoring systems. As such, the present system can conveniently be used when exercising.

[0062] Another objective of the present system is to employ the best placement for each of the ECG and PPG sensors on the body. With the present system, optimal placement of each of the ECG and PPG sensors on the body can be achieved, with the present system providing the required calibration.

[0063] Should two PPG devices T1, T2 be used as in FIGS. 3A and 3B, (i.e.: with one located on each of the patient's opposite hands or toes), then is then possible to determine if the blood wave from the heart reaches the two hands (or feet) at the same time. A detected time difference in the PWTT or pulse metrics for the arterial pulse seen in opposite hands could (for an adult) indicate a diabetic problem, or (for a newborn) indicate problems with a delayed sealing of the ductus arteriosis that occurs within 2-4 days after birth. Performing such a diagnostic procedure would inherently require that the two PPG devices be measuring the cardiac cycle at exactly the same time. In short, the two PPG devices would both need to be synchronized to the same cardiac cycle timing. As explained herein, the present system can be used to synchronize multiple devices to the same cardiac cycle.

[0064] Returning to FIG. 2, the PWTT TIME is the time from the calibration system C detecting the QRS wave to the time the target PPG device T1 (or T2) detects maximum arterial blood volume. This is represented on FIG. 2 as the signal from T1. (The signal from tracking device T2 is omitted from FIG. 2 for clarity).

[0065] As can be seen, the (cheaper, simpler) signaling device S will detect the QRS wave at a slightly delayed time as compared to the (more expensive and more accurate) calibration device C. Therefore, by adjusting targeting device T1's internal clock back by the time offset TO, a correct PWTT TIME can be determined. Stated another way, the difference in time between the signals from devices S and C will be provided to target device T1 enabling it to synchronize signaling to the cardiac cycle. Stated yet another way, after system calibration, in essence the signaling ECG system S shares the same internal clock of the calibration ECG system C.

[0066] An advantage of the present system is that it is only necessary to determine the timing of the QRS wave with each one of the S and C devices. Thus, it is only necessary to determine when the maximum PPG (and ECG) signals occurs. Importantly, it is not necessary to exactly determine the exact level of these signals. Therefore, an advantage of the present system is that different ECG and PPG systems can be used (with the present system compensating for differences i.e.: system calculation delays) between different manufacturers.

[0067] As shown in FIG. 3B, the calibration ECG device C can be removed after the time offset TO has been provided to the first PPG target device T1 (and optionally to a second target device T2). As a result, a single expensive calibration system C can be used to calibrate multiple target devices T1, T2, etc. This allows cheaper, fitness-monitoring watches and bands to be synchronized to a patient's cardiac cycle such that they can be used to accurately measure a fitness enthusiast's pulse metrics (after calculating PWTT). Periodic recalibration of the target device(s) can be done to the PPG device(s) T1 (and T2).

[0068] An important advantage of calibrating a fitness-monitoring watch or band (i.e.: signaling device S) to a patient's cardiac cycle is that the signaling device S can be a small, lightweight, inconspicuous and comfortable device that can be worn while exercising. As such, a more expensive, bulky, yet highly accurate ECG system (i.e.: calibration device C) need not be required during exercise or continued use.

[0069] To date, accurate PWTT measurements require an expensive, highly accurate ECG system that accurately detects the exact moment of the heart's QRS signal. This currently is done in a research setting, or as part of a clinical trial. The present calibration system avoids this problem. Using the present system and techniques, it is possible to accurately determine a patient's PWTT (since the inexpensive, less accurate ECG signaling system S is first accurately synchronized to the cardiac cycle. As a result, simple, cheaper ECG devices (such as those in various fitness watches and trackers) can be used to accurately determine PWTT. As such, the more expensive and accurate ECG calibration device C need only be used only for initial system calibration.

[0070] In optional embodiments, the leads of the signaling ECG system S can be disposed in opposite handlebars of an exercise machine. The leads of the signaling or calibration ECG devices can optionally be disposed in opposite sides or ends of a hand-held device (such as a smartphone or stethoscope).

[0071] FIGS. 4A and 4B illustrate an exemplary ECG calibration system for use with a signaling PPG system and target ECG system. In this example, an existing telemetry system 90 functions as the calibration system (i.e.: telemetry system 90 corresponds to system C in FIG. 3A). An arm strap 80 houses both the target and signaling systems (corresponding to illustrated T and S systems in FIGS. 3A and 3B). Arm strap system 80 wraps around the patient's arm (or leg) and includes a right electrode 54 and a PPG sensor 60. The left electrode 52 extends across the patient's chest to measure electrical signals on the left side of the patient's heart. In the embodiment shown in FIG. 4A, the present system simply piggy-back connects left chest electrode 52 on an existing telemetry system electrode 91 from telemetry system 90. Specifically, a stackable rivet-type electrode snap may be provided such that arm electrode 52 can quickly and easily be attached to the opposite chest electrode 91. In various embodiments, the electrodes may be wet or dry electrodes. An advantage of using wet electrodes is that they tend to provide a stronger, more stable signal. It is to be understood that the present system does not require telemetry system 90 to be used on the patient at the same time as the present system. As such, telemetry system 90 (and its associated electrode 91) can be removed from the patient after the calibration has been performed (as illustrated in FIG. 3B where C has been removed).

[0072] FIGS. 5A and 5B illustrate an exemplary adhesive patch system 150 housing both signaling and target devices (corresponding to systems S and T in FIGS. 3A and 3B), as follows. Integrated patch system 150 can be used to measure a person's PWTT/pulse waveform. Patch 150 may preferably comprise a left electrode 152 and a right electrode 154 for measuring ECG readings across the patient's heart. A PPG sensor 160 is also provided. Electrode 154 may also optionally be a “snap” electrode that simply piggy-back connects left chest electrode 152 on an existing telemetry system electrode (i.e.: electrode 91 from telemetry system 90 in FIG. 4A).

[0073] FIG. 5C illustrates a similar chest strap device 50 housing both signaling and target devices (again corresponding to systems S and T in FIGS. 3A and 3B), as follows. Chest belt or strap mounted device 50 can be used to measure a person's PWTT/pulse waveform. Strap device 50 may preferably comprise a strap body 51 with a left electrode 52 and a right electrode 54 for measuring ECG readings across the patient's heart. The PPG sensor 60 is disposed on the patient-facing side of strap body 51. FIG. 5C illustrates the positioning of the signaling device along the lines of the adhesive patch system of FIGS. 5A and 5B, but when the device is instead positioned within a chest strap worn by the patient.

[0074] FIGS. 6A and 6B illustrate yet another exemplary device housing both signaling and target devices (corresponding to systems S and T in FIGS. 3A and 3B), as follows. Device 10 is a chest or side mounted device for measuring pulse waveforms. Device 10 comprises a housing (that is preferably shaped to be hand-held, as shown), having a first (left chest) electrode 12 and a second (right chest) electrode 14 thereon. An ECG system (corresponding to signaling system S in FIGS. 3A and 3B) is disposed within the housing of device 10 and is in electrical communication with electrodes 12 and 14. When device 10 is positioned on a patient's chest as shown, electrodes 12 and 14 are thus positioned across the patient's heart to take ECG readings on the patient. At least one (but preferably a plurality) of PPG sensor(s) 20 (corresponding to target system T in FIGS. 3A and 3B) are also disposed on the housing of device 10. Logic for measuring the PPG signal from sensor 20 is disposed within the housing of device 10. Also disposed within the housing of device 10 are control and communication systems, and preferably a battery or other source of power. An optional right electrode lead can be plugged into the housing of device 10 such that the patient's ECG can be measured across the patient's torso (when the patient is in a prone position).

[0075] FIGS. 7A and 7B illustrate an exemplary handheld signaling and targeting device that is held in a patient's hands. Device 200 has a pair of electrode handles 202 onto which a user grasps. The user simply holds electrode handles 202 and then uses their thumb to push start button 203. The user then immediately moves their thumb or finger to be positioned against PPG sensor 204. Holding onto electrode handles 202 completes a circuit across the heart allowing the ECG system in device 200 to measure ECG waveforms. The PPG sensor 204 allows the PPG system in device 200 to measure PPG waveforms. As such, the ECG system in device 200 corresponds to the signaling system S and the PPG system in device 200 corresponds to the target system T in FIGS. 3A and 3B. FIG. 7B shows a bottom screen that optionally displays a Pulse Wave Transit Time.

[0076] As stated above, the present calibration system preferably uses a “composite” PPG signal that is synchronized to the cardiac cycle. As will now be explained, the composite PPG signal is preferably generated by comparing various lengths of PPG signal segments to one another, and these signal lengths are preferably segmented on the basis of repeating features in the cardiac cycle. As will also be explained, the generation of such a composite signal PPG waveform mitigates the current problems of signal noise and motion artifacts when measuring a patient's PPG signals, as follows.

[0077] When wearable sensors are used during exercise or other motion, the interference from internal and external sources such as from body movement, muscular contractions, wire friction, external RF interference, as well as noise from the sensor itself can make the raw sensor output appear unusable. Noise and signal errors are especially common when dealing with PPG sensors on ambulatory patients. Such errors and noise become very pronounced when a patient is moving. As such, accurate PPG monitoring has proven to be difficult to perform on people that are exercising or otherwise moving around.

[0078] Existing systems tend to use high and/or low pass filters in an attempt to clean up the data signal. Unfortunately, applying high and/or low pass filters to PPG data can, and often does, remove or change important underlying signal information, such as peaks and valleys in the time domain. Moreover, attempts at mathematically averaging the PPG data signals utilizing only its own waveform characteristics (onset, maximum, duration) have been similarly imperfect in removing spurious readings while providing consistent accurate results. As a result, the existing systems for determining pulse metrics required invasive procedures (such as right or left heart catheterization with pressure monitors) to measure cardiac pressures and pulsations. Such invasive tests are therefore only reserved for situations wherein the benefits of intervention are deemed to exceed the very significant risks of the tests themselves. In contrast, the present system instead provides is a system for quickly producing reliable PPG signals such that pulse metrics can be determined accurately (and preferably without having to restrain the motion of the patient, while also not using high or low pass filtering which can remove important data from the signal). As will be shown, this preferred system provides data from patients in motion that is consistently usable for analysis by removing large amounts of noise from motion and other artifacts.

[0079] In various aspects, the present invention provides systems for removing motion and ambient variability from PPG sensor data to improve discovery of the underlying unfiltered PPG waveform. As such, the present system's novel computerized logic system includes various optional circuitry and logic systems that remove or compensate for the effects of noise in the PPG signal, as follows.

[0080] As seen in FIG. 8A, an ECG signal 200 is measured over a plurality of cardiac cycles. Specifically, the onset of the heart's R-wave of the QRS complex occurs repeatedly at points 202. The PPG system measures the PPG signal 300's strength over the same time period (i.e.: over a plurality of cardiac cycles). In accordance with the present system, the PPG signal is then segmented in lengths corresponding to the length of the cardiac cycle. For example, a first segment 300.sub.1 will be measured over the first cardiac cycle (i.e. between times to and t.sub.1), a second segment 300.sub.2 will be measured over the second cardiac cycle (i.e. between times t.sub.1 and t.sub.2), and a third segment 300.sub.3 will be measured over the third cardiac cycle (i.e. between times t.sub.2 and t.sub.3). Next, as seen in FIG. 8B, segments 300.sub.1, 300.sub.2 and 300.sub.3 can be averaged to produce a representative or “composite” signal segment 300C. It is to be understood that the use of three segments in FIGS. 8A and 8B is merely exemplary. For example, a greater number of PPG signal segments 300n may be used to generate the representative or composite signal segment 300C.

[0081] As a result, the present system advantageously removes signal errors by taking a long PPG signal reading 300 (i.e.: lasting greater than several cardiac cycles), and then dividing the PPG signal into segments corresponding to the timing of the cardiac cycles. Preferably, the present PPG signal reading 300 is parsed or segmented based upon the timing of the cardiac QRS rhythm. Importantly, once the full waveform 300C has been generated and plotted in accordance with the present system, analysis of the exact shape of the waveform (or waveforms generated or extracted from this waveform) can be used to calculate various pulse wave metrics or observe other cardiac system features.

[0082] In other preferred aspects, generation of the composite PPG signal is performed by selecting PPG signal waveforms of similar R-to-R intervals of the pulse/cardiac cycle prior to the pulse/cardiac cycle in question. In various aspects, characteristics such as peak height, peak width, slope and duration can all contribute to a calculation that is accurately representative of that particular wave form.

[0083] FIG. 9 is a first side-by-side comparison of ECG and PPG signals showing steps in an alternate method of generating a PPG composite wave. Specifically, an ECG signal 400 is taken over five pulses/cardiac cycles (labelled pulses A, B, C, D and E). A PPG signal 500 is also taken over the same five pulses/ cardiac cycles (A, B, C, D and E). As can be seen, each of the five pulse lengths are not of exactly equal duration (which is to be expected as an individual's heart rate will vary over time). In accordance with this aspect of the present invention, the PPG signal will first be segmented and put into categories or “bins” representing segments of approximately equal lengths. Stated another way, all of the “short” segments can be grouped, categorized and analyzed together, all of the “intermediate length” segments can be categorized and analyzed together, and all of the “long duration” segments can be categorized and analyzed together. As such, separate composite waves can be generated for each of the short, intermediate and long categories of waveform segments. This is particularly useful in that some cardiac conditions or features may best be analyzed for an intermediate length pulse, whereas other cardiac conditions or features may best be analyzed for a short or long duration pulses. In fact, further insights may be gained by comparing composite PPG waveform segments from one category with another. For example, it is expected that certain cardiac conditions may be detected when a certain feature is seen in their intermediate length category while another feature is seen in their short or long composite waveform category. A wide variety of possibilities exist. It is therefore to be understood that a large variety of possible combinations exist, and the present invention is not limited to looking only at one category of waveform length or another. Instead, it is possible to compare composite waveforms from different categories to one another. Moreover, it is to be understood that the present invention is not limited to only three categories (i.e.: short duration, intermediate duration and long duration categories). For example, additional or fewer categories may be used (e.g.: very short, short, short-intermediate, standard-intermediate, long-intermediate, long and very long). It is to be understood that analysis may also be performed by only analyzing only one category's composite signal, or by also analyzing more than one category's composite signal.

[0084] In one exemplary approach seen in FIGS. 9 and 10, the pulses are sorted into categories based on the length of the previously measured pulse (and not the length of the current pulse being measured). The advantage of this novel approach is that it categorizes waveforms on the basis of similar ventricular fillings. Specifically, the filling stage of the heart in one cardiac cycle will correspond to the squeezing or emptying stage of the heart in the next cardiac cycle. Stated another way, the pre-contraction ventricular filling state will depend upon the time available to fill after the last contraction. Ventricular function will therefore vary beat to beat depending upon the variability of the pulse length. As such, R-to-R pulses are preferably compared in pairs, with the second pulse being categorized on the basis of the length of the first pulse, as follows.

[0085] In FIG. 9, Pulses B and E are categorized on the basis of their immediately previous pulses (i.e.: Pulse A and Pulse D). Since Pulses A and D were intermediate duration pulses; Pulses B and E are therefore placed together in the intermediate length category (even though Pulses B and E have considerably different lengths).

[0086] FIG. 10 shows a longer series of pulses A to I. The prior R-to-R categorization (in which two successive pulses are analyzed together) proceeds as follows. Pulses B and C are analyzed together as a first “2 beat complex”. C is placed into a category corresponding to the length of B. Next, pulses C and D are analyzed together as a second “2 beat complex”. D is placed into a category corresponding to the length of C. Next, pulses D and E are analyzed together and E is placed into a category corresponding to the length of D. Next, pulses E and F are analyzed together and F is placed into a category corresponding to the length of E, etc.

[0087] Moreover, it is to be understood that the categorizations of waveform segments illustrated in FIGS. 9 and 10 can also be re-categorized and then re-analyzed over a longer period of time. At that time, the various categories into which the segments are placed can also be changed to perform additional analysis. For example, it may make sense to record ECG signal 400 and PPG signal 500 over several hundred cardiac cycles. Once all this data has been recorded and the waveforms segmented into their R-to-R segments (or otherwise segmented based on repeating identifiable cardiac features), then the present system can go back and place the segments into different categories. For example, these segments can be placed into three categories (short, intermediate and long), or more categories (e.g.: very short, short, short-intermediate, standard-intermediate, long-intermediate, long and very long). It is to therefore be understood that an advantage of the present system is that different forms of analysis based on placing the same signal segments into different categories and then analyzing these segments at different times can yield different useful results. As can also be appreciated, not only does the present system therefore provide an excellent system of categorizing and analyzing waveforms, but it can also readily determine and report on a patient's heart rate variability.

[0088] It is to be understood that the present invention encompasses all forms of composite wave generation, and all forms of segmenting pulse waveforms to group the segments into self-similar groups, categories or bins of different time durations. For various cardiac conditions, analysis of one category (e.g.: the intermediate duration segments) may yield the best diagnostic results. For other cardiac conditions, analysis of another category (e.g.: the long duration segments) may yield the best diagnostic results. It all depends upon which medical condition the present system is diagnosing at the time. The advantage of the present system is that it provides a novel platform to categorize the waveform segments based on their relationships to the one another in general, and to the segment that immediately precedes it in particular.