`571-272-7822
`
`
`
`
`
`
`
`Paper 42
`Entered: December 6, 2022
`
`
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`____________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`____________
`
`APPLE, INC.,
`Petitioner,
`
`v.
`
`ALIVECOR, INC.,
`Patent Owner.
`____________
`
`IPR2021-00971
`Patent 10,595,731 B2
`___________
`
`
`
`
`Before ROBERT A. POLLOCK, ERIC C. JESCHKE, and
`DAVID COTTA, Administrative Patent Judges.
`
`POLLOCK, Administrative Patent Judge.
`
`
`
`
`JUDGMENT
`Final Written Decision
`Determining All Challenged Claims Unpatentable
`35 U.S.C. § 318(a)
`
`Denying In-Part and Dismissing In-Part as Moot
`Patent Owner’s Motion to Exclude Evidence
`37 C.F.R. § 42.64
`
`
`
`
`
`IPR2021-00971
`Patent 10,595,731 B2
`
`
`I.
`
`INTRODUCTION
`
`A. Background
`Apple, Inc. (“Petitioner”) filed a Petition for an inter partes review of
`claims 1–30 of U.S. Patent No. 10,595,731 B2 (Ex. 1001, “the ’731 patent”).
`Paper 2 (“Pet.”). AliveCor, Inc. (“Patent Owner”) timely filed a Preliminary
`Response. Paper 6 (“Prelim. Resp.”). Petitioner further filed an authorized
`Reply to the Preliminary Response (Paper 7); Patent Owner filed a
`responsive Sur-reply (Paper 8). Taking into account the arguments and
`evidence presented, we determined the information presented in the Petition
`established that there was a reasonable likelihood that Petitioner would
`prevail in demonstrating unpatentability of at least one challenged claim of
`the ’731 patent, and we instituted this inter partes review as to all challenged
`claims. Paper 10 (“DI”).
`After institution, Patent Owner filed a Patent Owner Response (Paper
`26, “PO Resp.”); Petitioner filed a Reply to the Patent Owner Response
`(Paper 29, “Reply”); Patent Owner filed a (corrected) Sur-reply (Paper 36,
`“PO Sur-reply”).
`Patent Owner also filed a motion to exclude (Paper 34, “Mot.”);
`Petitioner opposed the motion (Paper 36, “Opp. Mot.”); and Patent Owner
`filed a reply in support of its motion (Paper 38, “Reply Mot.”).
`An oral hearing was held on September 14, 2022, and a transcript of
`the hearing is included in the record. Paper 41 (“Tr.”).
`We have jurisdiction under 35 U.S.C. § 6. This decision is a Final
`Written Decision under 35 U.S.C. § 318(a) as to the patentability of claims
`1–30 of the ’731 patent. For the reasons discussed below, we hold that
`
`2
`
`
`
`IPR2021-00971
`Patent 10,595,731 B2
`
`Petitioner has demonstrated by a preponderance of the evidence that claims
`1–30 are unpatentable.
`
`B. Real Parties-in-Interest
`Petitioner identifies itself, Apple Inc., as the real party-in-interest. Pet.
`88. Patent Owner, identifies itself, AliveCor, Inc., as the real party-in-
`interest. Paper 6, 2.
`
`C. Related Matters
`According to Patent Owner:
`U.S. Patent No. 10,595,731 has been asserted by Patent
`Owner against Petitioner in AliveCor, Inc. v. Apple, Inc., Case
`No. 6:20-cv-01112-ADA, filed in the United States District
`Court for the Western District of Texas, and in Investigation
`No. 337-TA-1266 before the International Trade Commission,
`In the Matter of Certain Wearable Electronic Devices with
`ECG Functionality and Components Thereof. Apple also filed
`IPR petitions against the other patents asserted in those actions:
`IPR2021-00970 (USP 9,572,499) and IPR2021-00972 (USP
`10,638,941).
`Paper 6, 2; see Pet. 88. We further note that the ’731 patent at issue here is
`related by a chain of continuation applications to Application No.
`14/730,122, which issued as U.S. Patent No. 9,572,499 (“the ’499 patent”),
`challenged in IPR2021-00970. See Ex. 1001, code (63). As such, the ’731
`and ’499 patents share substantially the same specification.
`
`D. Priority Date of the ’731 Patent
`The ’731 patent claims priority to, inter alia, a series of provisional
`applications filed between December 12, 2013, and June 19, 2014. Ex. 1001,
`code (60); see Prelim. Resp. 4; Pet. 2 & nn. 1–3. Petitioner contends that the
`claims of the ’731 patent are not entitled the benefit of the earliest of those
`applications such that the critical date is March 14, 2014, the filing date of
`
`3
`
`
`
`IPR2021-00971
`Patent 10,595,731 B2
`
`provisional application No. 61/953,616. Pet. 2–3. Because Patent Owner
`does not contest this assertion, or the prior art status of any asserted
`reference, we need not determine whether the challenged claims are entitled
`to the benefit of the earliest filed provisional application. See generally
`Prelim. Resp. 4; PO Resp. 17, 19.
`
`E. Asserted Grounds of Unpatentability
`Petitioner asserts the following grounds of unpatentability (Pet. 1):
`
`Ground
`1
`
`2
`
`3
`
`4
`
`5
`
`Claims Challenged
`1, 7, 12, 13, 16, 17,
`23–26, 30
`1, 2, 4, 7, 12–14, 16–18,
`20, 23–26, 30
`3, 5, 6, 19, 21, 22
`
`8–11, 27–29
`
`15
`
`35 U.S.C §1 Reference(s)/Basis
`§ 103
`Shmueli2
`Shmueli, Osorio3
`
`§ 103
`
`§ 103
`
`§ 103
`
`§ 103
`
`Shmueli, Osorio,
`Li 20124
`Shmueli, Osorio,
`Kleiger5
`Shmueli, Osorio,
`Chan6
`
`
`1 The Leahy-Smith America Invents Act (“AIA”) included revisions to
`35 U.S.C. § 103 that became effective on March 16, 2013. Because we
`determine the priority date of the challenged claims is no earlier than the
`’731 patent’s filing date of March 14, 2014 (see infra), we apply the AIA
`versions of the statutory bases for unpatentability.
`2 WO2012/140559, publ. Oct. 18, 2012. Ex. 1004.
`3 U.S. 2014/0275840, publ. Sept. 18, 2014. Ex. 1005.
`4 Li Q, Clifford GD, “Signal quality and data fusion for false alarm
`reduction in the intensive care unit,” 45(6) J Electrocardiol. 596-603 (2012).
`(“Li” or “Li-2012”) Ex. 1006.
`5 Kleiger RE, Stein PK, “Bigger JT Jr. Heart rate variability: measurement
`and clinical utility.” 10(1) Ann Noninvasive Electrocardiol. 88–101 (2005).
`(“Kleiger”) Ex. 1033.
`6 U.S. Pat. No. 7,894,888, issued Feb. 22, 2011. Ex. 1048.
`
`4
`
`
`
`IPR2021-00971
`Patent 10,595,731 B2
`
`
`In support of its patentability challenge, Petitioner relies on, inter alia,
`the Declaration of Dr. Bernard R. Chaitman, M.D. Ex. 1003. Patent Owner
`similarly relies on the Declarations of Dr. Igor Efimov, Ph.D. Ex. 2001;
`Ex. 2016.
`
`F. The ’731 Patent and Relevant Background
`The ’731 patent relates to medical devices, systems, and methods for
`detecting cardiac conditions, including cardiac arrhythmias. Ex. 1001, 1:29–
`33, 2:17–25. In general:
`In response to the continuous measurement and recordation of
`the heart rate of the user, parameters such as heart rate (HR),
`heart rate variability (R-R variability or HRV), and heart rate
`turbulence (HRT) may be determined. These parameters and
`further parameters may be analyzed to detect and/or predict one
`or more of atrial fibrillation, tachycardia, bradycardia,
`bigeminy, trigeminy, or other cardiac conditions.
`Id. at 2:57–64; see id. at 18:52–63 (Table 2, listing atrial fibrillation, sinus
`and supraventricular tachycardias, bradycardia, bigeminy, and trigemini
`among the types of arrhythmias).
`According to Dr. Chaitman, “HRV analysis is an important tool in
`cardiology to help diagnose various types of arrhythmia.” Ex. 1003 ¶ 35.
`“HRV is defined as the variation of RR intervals with respect to time and
`reflects beat-to-beat heart rate (HR) variability,” and “can be accurately
`determined based on either ECG [electrocardiogram] data or PPG
`[photoplethysmography] data.” Id. ¶¶ 35–36. “An R-R interval represents a
`time elapsed between successive R-waves of a QRS complex7 of the ECG
`
`
`7 “[E]lectrical activity of the heart based on depolarization and repolarization
`of the atria and ventricles . . . typically show[s] up as five distinct waves on
`[an] ECG readout – P-wave, Q-wave, R-wave, S-wave, and T-wave.” Ex.
`1003 ¶ 29. “A QRS complex is a combination of the Q, R, and S waves
`
`5
`
`
`
`IPR2021-00971
`Patent 10,595,731 B2
`
`that occur between successive heart beats.” Id. ¶ 29. “If the RR intervals
`over a time period are close to each other in value, then ventricular rhythm is
`understood to be ‘regular.’ In contrast, if there are significant variations in
`the RR intervals over a time period, then the ventricular rhythm is
`understood to be ‘irregular.’” Id. ¶ 37 (citations omitted).
`The Specification explains that during cardiac arrhythmia, “the
`electrical activity of the heart is irregular or is faster (tachycardia) or slower
`(bradycardia) than normal,” and in some forms, “can cause cardiac arrest
`and even sudden cardiac death.” Ex. 1001, 1:40–44. The ’731 patent
`identifies atrial fibrillation as the most common form of cardiac
`arrhythmia—which occurs when electrical conduction through the atria of
`the heart is irregular, fast, and disorganized, leading to irregular activation of
`ventricles. Id. at 1:44–49. Although atrial fibrillation may cause no
`symptoms, it is associated with palpitations, shortness of breath, fainting,
`chest pain, congestive heart failure, as well as atrial clot formation, which
`can lead to clot migration and stroke. Id. at 1:44–51. “Atrial fibrillation is
`typically diagnosed by taking an electrocardiogram (ECG) of a subject,
`which shows a characteristic atrial fibrillation waveform.” Id. at 1:52–54.
`The Specification discloses body-worn devices for detecting the
`occurrence of arrhythmias using a combination of ECG and PPG electrodes.
`See, e.g., claim 1. PPG, or photoplethysmography, uses an optical sensor to
`detect the fluctuation of blood flow, and can provide a measure of heart rate.
`Id. at 25:21–24. According to the Specification, fluctuations in heart rate not
`explained by changing activity levels may be interpreted as an advisory
`
`
`occurring in succession and represents the electrical impulse of a heartbeat
`as it spreads through the ventricles during ventricular depolarization.” Id.
`
`6
`
`
`
`IPR2021-00971
`Patent 10,595,731 B2
`
`condition for recording an ECG, or electrocardiogram, which is a typical
`method for diagnosing episodes of arrhythmia. Id. at 1:52–54, 1:60–65,
`25:1–35.
`The collected data may also be analyzed using machine learning
`algorithms to, for example, determine appropriate trigger thresholds, detect
`and predict health conditions, or provide a heart health score. See, e.g., id. at
`3:43–4:16, 8:38–41, 9:8–11, 12:44–64. “The machine learning based
`algorithm(s) may allow software application(s) to identify patterns and/or
`features of the R-R interval data and/or the raw heart rate signals or data to
`predict and/or detect atrial fibrillation or other arrhythmias.” Id. at 9:8–11. In
`particular,
`[a]ny number of machine learning algorithms or methods may
`be trained to identify atrial fibrillation or other conditions such
`as arrhythmias. These may include the use of decision tree
`learning such as with a random forest, association rule learning,
`artificial neural network, inductive logic programming, support
`vector machines, clustering, Bayesian networks, reinforcement
`learning, representation learning similarity and metric learning,
`sparse dictionary learning, or the like.
`Id. at 9:66–10:9.
`Figure 14, reproduced below, shows one embodiment of a body-worn
`device. Id. at 6:21–23.
`
`7
`
`
`
`IPR2021-00971
`Patent 10,595,731 B2
`
`
`Figure 14, shows “smart watch 1400 which includes at least one heart
`rate monitor 1402 and at least one activity monitor 1404,” such as an
`accelerometer. Id. at 24:66–25:1, 25:13–30. Analysis of signals from these
`monitors can be used to “determine if heart rate and activity measurements
`represent an advisory condition for recording an ECG,” and trigger signals
`for recording an ECG if an advisory condition is detected. Id. at 25:1–12.
`Figure 10, illustrated below, shows another embodiment involving a
`body-worn device. Id. at 6:3–5.
`
`Figure 10 illustrates “a method for monitoring a subject to determine when
`to record an electrocardiogram (ECG).” Id. at 23:20–22. According to the
`Specification:
`In FIG. 10, a subject is wearing a continuous heart rate monitor
`(configured as a watch 1010, including electrodes 1016), shown
`in step 1002. The heart rate monitor transmits (wirelessly 1012)
`heart rate information that is received by the smartphone 1018,
`as shown in step 1004. The smartphone includes a processor
`that may analyze the heart rate information 1004, and when an
`irregularity is determined, may indicate 1006 to the subject that
`an ECG should be recorded.
`
`8
`
`
`
`IPR2021-00971
`Patent 10,595,731 B2
`
`Id. at 23:22–30. In some embodiments, the ECG device is “present in
`a smart watch band or a smart phone.” Id. at 25:36–37. “The ECG,
`heart rate, and rhythm information can be displayed on the computer
`or smartphone, stored locally for later retrieval, and/or transmitted in
`real-time to a web server.” Id. at 25:48–50.
`
`G. Challenged Claims
`Petitioner challenges claims 1–30, of which claims 1, 17, and 25 are
`independent. Of these, claim 1 recites:
`1. A smart watch to detect the presence of an arrhythmia of a
`user, comprising:
`a processing device;
`a photoplethysmography (“PPG”) sensor operatively coupled to
`the processing device;
`an ECG sensor, comprising two or more ECG electrodes, the
`ECG sensor operatively coupled to the processing device;
`a display operatively coupled to the processing device; and
`a memory, operatively coupled to the processing device, the
`memory having instructions stored thereon that, when executed
`by the processing device, cause the processing device to:
`receive PPG data from the PPG sensor;
`detect, based on the PPG data, the presence of an
`arrhythmia;
`receive ECG data from the ECG sensor; and
`confirm the presence of the arrhythmia based on the ECG
`data.
`Independent claims 17 and 25 recite similar limitations but are respectively
`drawn to “[a] method to detect the presence of an arrhythmia of a user on a
`smart watch,” and “non-transitory computer-readable storage medium
`including instructions.”
`
`9
`
`
`
`IPR2021-00971
`Patent 10,595,731 B2
`
`Among the dependent claims, claims 2, 14, and 18 relate to the use of
`
`motion sensor (inertial) data; claims 4 and 20 relate to “determin[ing]
`heartrate variability (‘HRV’) data from the PPG data, and detect[ing], based
`on the HRV data, the presence of the arrhythmia”; claims 3, 5, 6, 19, 21, and
`22 recite “a machine learning algorithm trained to detect arrhythmias”; and
`claim 15 recites a device “configured to display an ECG rhythm strip for the
`ECG data.”
`
`H. Overview of the Asserted References
`1) Shmueli (Exhibit 1004)
`Shmueli, titled “Pulse Oximetry Measurement Triggering ECG
`Measurement,” addresses “solutions . . . for monitoring infrequent events of
`irregular ECG.” Ex. 1004, 2.8 According to Shmueli, “[t]he present
`invention preferably performs measurements of intermittent irregular heart-
`related events without requiring the fixed wiring of the ECG device to the
`patient.” Id. at 8.
`Shmueli discloses body-worn cardiac monitoring devices “equipped
`with two types of sensing devices: an oximetry (SpO2) measuring unit and
`an ECG measuring unit.” Id.9 Shmueli’s Figures 1A, 1B, and 4, reproduced
`below, exemplify one embodiment (annotations by Petitioner in red):
`
`
`8 Throughout this opinion, we cite to the native pagination. For clarity with
`respect to citations to Shmueli, we understand the native pagination to be the
`numbers at the top of the page.
`9 As used by Shmueli “the terms ‘oxygen saturation in the blood’, ‘blood
`oxygen saturation’, ‘pulse oximeter’, oximetry, SpO2, and
`photoplethysmography have the same meaning and may be used
`interchangeably, except for those places where a difference between such
`terms is described.” Id. at 7; see Tr. 6:22–7:12, 73:18–21, 95:7–11.
`
`10
`
`
`
`IPR2021-00971
`Patent 10,595,731 B2
`
`
`Pet. 12. Figures 1A, 1B, and 3 show three views of a wrist-mount heart
`monitoring device having three ECG electrodes 14 and a PPG sensor 13.
`Ex. 1004, 6, 9–10. Figure 1A shows two of the ECG electrodes, 14/16, on
`the face of the device. Id. at 9. Figure 1B shows a third ECG electrode,
`14/15, along with PPG sensor 13, of the back of the device. Id. Figure 3
`shows the device as worn on a patient’s wrist, with PPG sensor 13 and ECG
`electrode 14/15 in contact with the patient’s left wrist and ECG electrodes
`14/16 in contact with two fingers of the patient’s right hand. Id. Petitioner
`annotates each of Figures 1A, 1B, and 3 with arrows identifying the ECG
`electrodes. Petitioner has also annotated Figure 1B with an arrow identifying
`PPG sensor 13. In connection with these devices, Shmueli discloses
`a method for triggering measurement of electrocardiogram
`(ECG) signal of a subject, the method including the steps of:
`continuously measuring SpO2 at least one of a wrist and a
`finger of the subject, detecting an irregular heart condition from
`the SpO2 measurement, notifying the subject to perform an
`ECG measurement, and initiating ECG measurement at least
`partially at the wrist.
`Id. at 2; see Abstract.
`Shmueli explains that “[d]eriving heart beat rate from oximetry, as
`well as other artifacts of the heart activity and blood flow, is . . . known in
`
`11
`
`
`
`IPR2021-00971
`Patent 10,595,731 B2
`
`the art,” as are various body-worn oximetry devices. Id. at 8. Shmueli further
`explains that the use of oximetry in combination with ECG measurements is
`also known in the art. Id. Shmueli states, for example, that “US patent No.
`7,598,878 (Goldreich) describes a wrist mounted device equipped with an
`ECG measuring device and a SpO2 measuring device.” Id. However,
`Shmueli, notes “Goldreich does not teach interrelated measurements of ECG
`and SpO2” and, thus, does not “enable a patient to perform ECG
`measurement as soon as an irregular heart activity develops and without
`requiring the ECG to be constantly wired to the patient.” Id. According to
`Shmueli:
`The present invention resolves this problem by providing a
`combined oximetry and electrocardiogram measuring system
`and a method in which the oximetry measurement is performed
`continuously and/or repeatedly, and the ECG measurement is
`triggered upon detection of an intermittent irregular heart-
`related events without requiring the fixed wiring of the ECG
`device to the patient.
`Id. Consistent with this disclosure, Shmueli claims:
`1. A method for triggering measurement of electrocardiogram
`(ECG) signal of a subject, the method comprising the steps of:
`continuously measuring SpO2 at least one of a wrist and a
`finger of said subject;
`detecting an irregular heart condition from said SpO2
`measurement;
`notifying said subject to perform an ECG measurement;
`and
`initiating ECG measurement at least partially at said wrist.
`Id. at 16.
`
`12
`
`
`
`IPR2021-00971
`Patent 10,595,731 B2
`
`
`
`Shmueli Figure 7 is reproduced below:
`
`“Fig. 7 is a simplified flow chart of a software program preferably executed
`by the processor of the wrist-mounted heart monitoring device.” Id. at 7; see
`also id. at 12–13 (further describing the steps of the software program
`illustrated in Figure 7).
`
`2) Osorio (Exhibit 1005)
`Osorio, titled “Pathological State Detection Using Dynamically
`Determined Body Data Variability Range Values,” “relates to medical
`device systems and methods capable of detecting a pathological body state
`of a patient, which may include epileptic seizures, and responding to the
`same.” Ex. 1005 ¶ 2. Although broadly referencing “a pathological body
`state,” Osorio repeatedly exemplifies such conditions in terms of detecting
`epileptic events. See, e.g., id. ¶ 37 (referencing values that may “be
`indicative of a certain pathological state (e.g., epileptic seizure)”), ¶ 46 (“In
`
`13
`
`
`
`IPR2021-00971
`Patent 10,595,731 B2
`
`one embodiment, the pathological state is an epileptic event, e.g., an
`epileptic seizure.”), ¶ 56 (“HRV range may be taken as an indication of an
`occurrence of a pathological state, e.g., an epileptic seizure”), ¶ 66 (“The
`dynamic relationship between non-pathological HRVs and activity levels
`may be exploited to detect pathological states such as epileptic seizures”).
`Consistent with the broad disclosure and narrow exemplification in
`the body of its specification, Osorio’s claim 1 is directed to “[a] method for
`detecting a pathological body state of a patient,” whereas claim 7 limits the
`pathological state to an epileptic event. Id. at claim 1, claim 7; also compare
`id. at claim 14, with claim 17 (similarly limiting a pathological state to an
`epileptic event).
`According to Osorio, the disclosed methods, systems, and related
`devices, detect a pathological state of a patient by determining when a body
`data variability value, or “BDV,” is outside of a “value range,” and where
`the threshold levels of that range vary in response to the patient’s physical
`activity (measured by, e.g., an accelerometer) or mental/emotional state. See,
`e.g., id. at code (57), ¶¶ 3–8, 28, 33, 35. In this respect, Osorio states that
`“false negative and false positive detections of pathological events may be
`reduced by dynamically determining pathological or non-pathological ranges
`for particular body indices based on activity type and level or other variables
`(e.g., environmental conditions).” Id. ¶ 36.
`
`14
`
`
`
`IPR2021-00971
`Patent 10,595,731 B2
`
`
`Osorio’s Figure 1 is reproduced below.
`
`Figure 1 shows a schematic representation of medical device system
`100, including kinetic sensor(s) 212 and body signal sensor(s) 282 connected
`to medical device 200 by leads 211 and 281, respectively. Id. ¶ 33.
`“[A]ctivity sensor(s) 212 may each be configured to collect at least one
`signal from a patient relating to an activity level of the patient,” and include,
`for example, an accelerometer, an inclinometer, a gyroscope, or an
`ergometer. Id. Figure 1 also shows a current body data variability (BDV)
`module 265, which may “may comprise an O2 saturation variability (O2SV)
`module 330 configured to determine O2SV from O2 saturation data,” and
`
`15
`
`
`
`IPR2021-00971
`Patent 10,595,731 B2
`
`“an HRV module 310 configured to determine HRV from heart rate data.”
`Id. ¶¶ 10, 13, 53, Fig. 2C. Osorio discloses that “medical device system 100
`may be fully or partially implanted, or alternatively may be fully external.”
`Id. ¶ 33.
`Figure 8, reproduced below, shows one embodiment of Osorio’s
`monitoring method.
`
`Figure 8 shows that an activity level is determined at 810, and a non-
`pathological BDV range is determined at 820 based on the activity level. Id.
`¶ 77. A current BDV is determined at 840 and compared to the non-
`pathological BDV range at 850. Id. ¶ 78. If the current BDV is outside the
`non-pathological range, then a pathological state is determined at 860 and a
`further action, such as warning, treating, or logging the occurrence and/or
`severity of the pathological state, is taken at 870. Id.
`
`16
`
`
`
`IPR2021-00971
`Patent 10,595,731 B2
`
`According to Osorio, body indices that may be the subject of BDV
`
`monitoring include:
`heart rhythm variability, a heart rate variability (HRV), a
`respiratory rate variability (RRV), a blood pressure variability
`(BPV), a respiratory rhythm variability, respiratory sinus
`arrhythmia, end tidal CO2 concentration variability, power
`variability at a certain neurological index frequency band (e.g.,
`beta), an EKG morphology variability, a heart rate pattern
`variability, an electrodermal variability (e.g., a skin resistivity
`variability or a skin conductivity variability), a pupillary
`diameter variability, a blood oxygen saturation variability, a
`kinetic activity variability, a cognitive activity variability,
`arterial pH variability, venous pH variability, arterial-venous
`pH difference variability, a lactic acid concentration variability,
`a cortisol level variability, or a catecholamine level variability.
`Id. ¶ 43; see also id. ¶ 42 (similar) ¶¶ 45–46 (monitoring heart rate for
`episodes of tachycardia and bradycardia). “In one embodiment, the severity
`[of a pathological state] may be measured by a magnitude and/or duration of
`a pathological state such as a seizure, a type of autonomic change associated
`with the pathological state (e.g., changes in heart rate, breathing rate, brain
`electrical activity, the emergence of one or more cardiac arrhythmias, etc.).”
`Id. ¶ 71.
`
`With respect to HRV, in particular, Osorio teaches: “By monitoring
`the patient’s activity level, HR, and HRV, it is possible to determine when
`the patient’s HRV falls outside the non-pathological ranges as the patient’s
`activity levels change over time.” Id. ¶ 66. Osorio’s Figure 4A, reproduced
`below, shows heart rate variability as a function of activity level. See id.
`¶ 58.
`
`17
`
`
`
`IPR2021-00971
`Patent 10,595,731 B2
`
`
`Figure 4A plots a patient’s heart rate (HR) on the Y-axis and a
`patient’s activity level on the X-axis. Id. Markers A1 though A4 represent
`increasing activity from a sleep state (A1) through vigorous activity (A4). Id.
`Boundary lines 410 and 420, respectively, represent the upper and lower
`limits of non-pathological heart rate, and include representative ranges R1
`through R4. Id. at Fig. 4A. According to Osorio,
`the upper and lower bounds of the non-ictal[10] HR region
`increase as activity level increases (e.g., from a sleep state to a
`resting, awake state) and reach their highest values for
`strenuous exertion. In addition, the width of the non-
`pathological HR ranges narrows as activity levels and heart
`rates increase, which is consistent with the known reduction in
`HRV at high levels of exertion. When the patient is in a non-
`pathological state (e.g., when an epileptic patient is not having a
`seizure), for a particular activity level the patient’s HRV should
`
`
`10 “Ictal” refers to the active, middle stage of a seizure and corresponds with
`intense electrical brain activity. See https://epilepsyfoundation.org.au/
`understanding-epilepsy/seizures/seizure-phases/.
`
`18
`
`
`
`IPR2021-00971
`Patent 10,595,731 B2
`
`
`fall within a non-pathological HRV range associated with that
`activity level.
`Id. ¶ 58.
`Osorio further presents Figure 11 as “depict[ing] pathological and
`non-pathological BDV (e.g., HRV) value ranges.” Id. ¶¶ 23, 91. In this
`illustration, Osorio shows that HRV values falling below 0.5 bpm and above
`4 bpm are always pathological when activity level is low (e.g., resting or
`walking), whereas intermediate HRV values (0.5–4 bpm) may be
`pathological when considered in light of the patient’s activity level. Id.
`Osorio further notes that the boundaries between normal and pathological
`may be adjusted based on an individual’s physiology. “For example, in an
`epilepsy patient also suffering from tachycardia, and having base resting
`heart rate of 100-110 bpm, a decline in heart rate to 70 bpm may be
`indicative of a seizure slowing down the heart rate, even though a heart rate
`of 70 bpm is generally ‘normal’ across a typical population.” Id. ¶ 45.
`
`Kleiger (Exhibit 1033)
`3)
`Kleiger is a review article regarding the measurement and clinical
`utility of heart rate variability (HRV). Ex. 1033, Title. Kleiger discloses
`various methods for quantifying HRV including time domain, spectral or
`frequency domain, geometric, and nonlinear methods. Id. at 88. According
`to Kleiger:
`The greatest variation of heart rate occurs with circadian
`changes, particularly the difference between night and day heart
`rate, mediated by complex and poorly understood
`neurohormonal rhythms. Exercise and emotion also have
`profound effects on heart rate. Fluctuations in heart rate reflect
`autonomic modulation and have prognostic significance in
`pathological states.
`Id. (internal citation numbers omitted).
`
`19
`
`
`
`IPR2021-00971
`Patent 10,595,731 B2
`
`
`Long-term, usually 24-hour recordings, can be used to assess
`autonomic nervous responses during normal daily activities in
`health, disease, and in response to therapeutic interventions,
`e.g., exercise or drugs. RR interval variability is useful for
`assessing risk of cardiovascular death or arrhythmic events,
`especially when combined with other tests, e.g., left ventricular
`ejection fraction or ventricular arrhythmias.
`Id. at Abstract.
`
`4) Li 2012 (Exhibit 1006)
`Li 2012 investigates algorithms for reducing cardiac monitor false
`alarms (“FA”) in an intensive care setting. Ex. 1006, 1. Li 2012 explains that
`a lack of integration between different sensors results in frequent false
`alarms in intensive care units. Id. at Abstract. To reduce these false alarms,
`Li 2012
`present[s] a novel framework for FA reduction using a machine
`learning approach to combine up to 114 signal quality and
`physiological features extracted from the electrocardiogram,
`photoplethysmograph, and optionally the arterial blood pressure
`waveform. A machine learning algorithm was trained and
`evaluated on a database of 4107 expert-labeled life-threatening
`arrhythmias, from 182 separate ICU visits.
`Id. According to Li 2012, the resulting algorithm reduced false alarms with
`without substantial suppression of true alarms. Id. at Abstract, 7, Table 6.
`For example, “[f]or the ventricular tachycardia alarms, the best FA [false
`alarm] suppression performance was 30.5% with a TA [true alarm]
`suppression rate below 1%.” Id. at Abstract.
`
`5) Chan (Exhibit 1048)
`Chan discloses:
`A wristwatch worn by a user for measuring a three-lead ECG
`[that] includes three electrodes placed separately on the front,
`either side, and back or strap thereof. The wristwatch further
`
`20
`
`
`
`IPR2021-00971
`Patent 10,595,731 B2
`
`
`includes an electrode panel having the electrode on the front or
`either side of the watch, sensing elements, pressure, infrared or
`impedance detectors, and circuits. The electrode panel is
`capable of sensing the contact or press of fingers to trigger the
`ECG measuring. While the electrode in the back-side of the
`watch contacts the hand wearing the watch, the electrode and
`electrode panel on the front or either side of the watch are
`pressed by fingers from the other hand, and the electrode in the
`strap contacts the abdomen or left leg simultaneously. Thus, a
`three-lead ECG can be measured. ECG data can be transmitted
`to a personal or hospital computer by wireless networks or flash
`memory.
`Ex. 1048, Abstract.
`Chan’s Figures 1A and 1B, reproduced below, show an embodiment
`of the disclosed three-lead ECG wristwatch.
`
`Figures 1A and 1B, respectively, show the front and rear of a three-lead
`ECG wristwatch. Id. at 2:21–22. Figure 1A shows ECG electrode 4, sensing
`element 6 (which can detect “pressure, impedance or infrared for
`recognizing the contact or press made by fingers to initiate an ECG
`
`21
`
`
`
`IPR2021-00971
`Patent 10,595,731 B2
`
`measurement”), and display 7, which may be an LCD. Id. at 2:44–56.
`Display 7 can display text (e.g., time, heart rate, and, condition (normal vs
`arrhythmia) as well as “graph/animation, for an event reminding 13 and
`ECG waveforms 14.” Id. at 2:56–59; see also id. at 4:56–59 (stating, with
`reference to Figure 7, that “display 57 can show users time, heart rate,
`waveforms and any other information 61, such as activity level and
`temperature, if needed”).
`
`Chan Figure 2 is reproduced below.
`
`Figure 2 shows an embodiment of the three-lead ECG watch having a third
`lead 5 on the strap 11. Id. at 2:24–25, 3:1–4.
`
`22
`
`
`
`IPR2021-00971
`Patent 10,595,731 B2
`
`
`Chan Figure 3B is reproduced below.
`
`Figure 3B “demonstrate[s] how to place the wristwatch to make electrodes
`be contacted by both hands.” Id. at 2:26–28, 3:5–22.
`
`II. ANALYSIS
`
`A. Legal Standards
`“In an IPR, the petitioner has the burden from the onset to show with
`particularity why the patent it challenges is unpatentable.” Harmonic Inc. v.
`Avid Technology, Inc., 815 F.3d 1356, 1363 (citing 35 U.S.C. § 312(a)(3)
`(requiring inter partes review petitions to identify “with particularity . . . the
`evidence that supports the grounds for the challenge to each claim”)). This
`burden of persuasion never shifts to Patent Owner. See Dynamic Drinkware,
`LLC v. Nat’l Graphics, Inc., 800 F.3d 1375, 1378 (Fed. Cir. 2015)
`(discussing the burden of proof in inter partes review).
`In KSR International Co. v. Teleflex Inc., 550 U.S. 398 (2007), the
`Supreme Court reaffirmed the framework for determining obviousness set
`
`23
`
`
`
`IPR2021-00971
`Patent 10,595,731 B2
`
`forth in Graham v. John Deere Co., 383 U.S. 1 (1966). The KSR Court
`summarized the four factual inquiries set forth in Graham (383 U.S. at 17–
`18) that are appli