`
`IPR2021-00970
`
`U.S. Patent No. 9,572,499
`PATENT OWNER’S SUR-REPLY
`
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`________________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`________________
`
`APPLE, INC.,
`Petitioner,
`
`v.
`
`ALIVECOR, INC.,
`Patent Owner
`________________
`IPR2021-00970
`U.S. Patent No. 9,572,499
`________________
`
`PATENT OWNER’S SUR-REPLY
`
`
`
`
`
`
`
`

`

`
`
`IPR2021-00970
`
`U.S. Patent No. 9,572,499
`PATENT OWNER’S SUR-REPLY
`
`
`TABLE OF CONTENTS
`
`B. 
`
`C. 
`
`Page
`INTRODUCTION ........................................................................................... 1 
`I. 
`THE BOARD SHOULD REJECT APPLE’S NEW EVIDENCE .................. 2 
`II. 
`III.  THE PRIOR ART DOES NOT RENDER ARRHYTHMIA
`DETECTION OBVIOUS ................................................................................ 5 
`A.  AliveCor’s Expert Did Not Agree That The Prior Art Teaches
`Or Suggests Arrhythmia Detection ....................................................... 6 
`Dr. Chaitman’s Shmueli Admissions Were Not Limited To
`Embodiments ......................................................................................... 7 
`Apple Mischaracterizes AliveCor’s Arguments And Shmueli’s
`Disclosures ............................................................................................ 8 
`Furthermore, Apple’s Positions Are Driven By Hindsight ................... 9 
`Shmueli’s Disclosures Teach A POSITA That Its System Is
`Directed To Heart Attacks, Not Arrhythmias ..................................... 10 
`1. 
`There Is No Record Evidence To Contradict Dr. Efimov
`That “Irregular Heart Condition” Refers To Heart
`Attacks ...................................................................................... 11 
`Apple Mischaracterizes AliveCor’s SpO2 Detection
`Arguments And Argues Against Yet Another Strawman......... 12 
`Shmueli’s CO2 Detection Disclosures Plainly indicate to
`a POSITA that it is Directed To Heart Attacks ........................ 13 
`Osorio Is Directed To Neurological Conditions, Not
`Arrhythmia .......................................................................................... 14 
`IV.  A POSITA WOULD NOT HAVE BEEN MOTIVATED TO
`COMBINE SHMUELI AND OSORIO ........................................................ 16 
`V.  NONE OF THE CITED ART RENDERS OBVIOUS MACHINE
`LEARNING ALONE OR IN COMBINATION ........................................... 17 
`1. 
`The Machine Learning Claims Limit The Detection Step,
`Which Is Performed By PPG .................................................... 20 
`
`D. 
`E. 
`
`F. 
`
`2. 
`
`3. 
`
`
`
`i
`
`

`

`
`
`IPR2021-00970
`
`U.S. Patent No. 9,572,499
`PATENT OWNER’S SUR-REPLY
`
`
`
`
`2. 
`
`3. 
`
`Hu 1997 Does Not Teach Application Of Machine
`Learning To PPG Data .............................................................. 21 
`Shmueli’s “Detection Parameters” Is Not Machine
`Learning .................................................................................... 22 
`VI.  ALIVECOR’S POSITA DEFINITION IS SUPPORTED BY THE
`RECORD EVIDENCE AND SHOULD BE ADOPTED ............................. 22 
`VII.  CONCLUSION .............................................................................................. 24 
`
`
`
`
`
`ii
`
`

`

`
`
`IPR2021-00970
`
`U.S. Patent No. 9,572,499
`PATENT OWNER’S SUR-REPLY
`
`
`INTRODUCTION
`Apple’s Reply asks this Board for a second bite at its Petition, abandoning its
`
`I.
`
`Petition positions by recasting its arguments and improperly submitting new
`
`evidence. This new evidence includes both hearsay testimony from Dr. Stultz (its
`
`ITC expert), as well as unauthenticated prior art references and other testimony cited
`
`for the first time. Importantly, Dr. Stultz did not submit a declaration in this IPR.
`
`Yet, Apple nevertheless heavily cites to Dr. Shultz’s out-of-court testimony in an
`
`apparent attempt to rehabilitate its IPR expert (Dr. Chaitman), who is barely cited at
`
`all in its Reply. Apple’s use of Dr. Stultz’s testimony is not only an improper attempt
`
`to submit new evidence, it is impermissible hearsay, and AliveCor was not given an
`
`opportunity to cross-examine him in this forum. His testimony, and any argument
`
`based on that testimony, should be struck on that basis alone. So too should the other
`
`additional new evidence submitted in Reply, which includes almost two dozen newly
`
`added prior art references, expert reports from other proceedings, and testimony
`
`transcripts, none of which was cited in the Petition. Just like Apple’s new expert
`
`testimony, this new evidence was all improperly submitted for the first time on
`
`Reply, and the record is devoid of any argument or evidence that any of it is authentic
`
`or that they were in fact printed publications. It should all be struck, as should any
`
`arguments based on that evidence.
`
`
`
`1
`
`

`

`IPR2021-00970
`
`U.S. Patent No. 9,572,499
`PATENT OWNER’S SUR-REPLY
`
`Even if that evidence is not struck, it fails to establish unpatentability.
`
`
`
`Tellingly, in many instances Apple does not address AliveCor’s arguments head on,
`
`instead reading in teachings from the prior art’s general disclosures. Apple would
`
`have this Board find that Shmueli’s teaching of the broad genus of irregular heart
`
`conditions and Osorio’s teaching of the broad genus of pathological conditions are
`
`both necessarily teachings of the species of arrhythmia. Yet Apple concludes so
`
`without applying the proper legal framework that the prior art disclosures must be
`
`viewed in context to ascertain what they teach to a POSITA. And as AliveCor
`
`established in its Response, when the references are reviewed by the skilled artisan
`
`in their entirety, it is clear that Shmueli is directed to techniques for detecting and
`
`preventing heart attacks, while Osorio is directed to neurological conditions such as
`
`seizures. The references’ disclosures are all in service of these broader contextual
`
`purposes. Indeed, perhaps recognizing that the primary prior art references are
`
`directed at non-analogous technical fields, Apple recasts its obviousness argument
`
`to rely on Amano, a reference that is not part of the grounds.
`
`II. THE BOARD SHOULD REJECT APPLE’S NEW EVIDENCE
`Apple submits almost two dozen new prior-art exhibits in its Reply, including
`
`testimony from its ITC expert (Dr. Stultz) and multiple unauthenticated printed
`
`publications. Both categories of new evidence should be rejected. With respect to
`
`
`
`2
`
`

`

`IPR2021-00970
`
`U.S. Patent No. 9,572,499
`PATENT OWNER’S SUR-REPLY
`
`the new Stultz testimony—Exhibits 1081 (ITC expert report), 1082 (ITC
`
`
`
`deposition), 1083 (ITC demonstratives), and 1072-1073 (ITC hearing testimony)—
`
`it is untimely and submitted for the first time on Reply. Even if it was not untimely,
`
`it should be rejected, it is hearsay. Yet despite these failings, Apple bases many of
`
`its new Reply obviousness arguments on Dr. Stultz’s testimony. While these
`
`arguments are substantively incorrect as discussed in more detail below, the Board
`
`should give them and Dr. Stultz’s testimony zero weight as based on inadmissible,
`
`untimely hearsay evidence. Fundamental fairness likewise demands that the
`
`testimony be ignored, as Apple should have included this testimony in the Petition,
`
`giving both the Board and Patent Owner the full opportunity to consider and rebut.1
`
`Also problematic is Apple’s introduction of new “secondary” references to
`
`bolster Shmueli’s lacking disclosure. In Section II.A.3 of the Reply, titled
`
`“Secondary Evidence Also Reinforces The Petition,” Apple introduces additional,
`
`new secondary references to support the proposition that “irregular activity,” as that
`
`phrase is used in Shmueli, must refer to arrhythmia. See Reply at 8-9 (citing Exs.
`
`1061, 1062, 1066, 1067, 1068). This evidence is untimely, should have been
`
`
`1 As Apple is well-aware, the PTAB rules do not permit expert declarations on
`
`Surreply.
`
`
`
`3
`
`

`

`IPR2021-00970
`
`U.S. Patent No. 9,572,499
`PATENT OWNER’S SUR-REPLY
`
`submitted with the Petition, and constitutes a material change in obviousness
`
`
`
`positions meriting a finding of nonobviousness. Shifting positions in reply is
`
`prohibited by both the Trial Practice Guide, statute, and controlling Federal Circuit
`
`precedent. See Wasica Finance GmbH v. Continental Automotive Sys., Inc., 853
`
`F.3d 1272, 1286-87 (Fed. Cir. 2017) (citing 35 U.S.C. § 312(a)(3) (requiring
`
`petitions to identify “with particularity ... the grounds on which the challenge to each
`
`claim is based”); Trial Practice Guide, 77 F. Reg. 48,756, 48,767 (Aug. 14, 2012)
`
`(prohibiting parties from submitting evidence necessary for a prima facie showing
`
`of obviousness in a reply)). This admittedly “secondary” evidence is plainly new
`
`and could have been cited in the Petition if necessary to support Apple’s prima facie
`
`obviousness positions.
`
`In addition to being improperly cited for the first time in Reply, most of
`
`Apple’s reply evidence is hearsay—Exhibits 1065-1068, 1074-1080 and 1085 are
`
`all alleged printed publications, which are being relied upon for the truth of the
`
`matter asserted. They are hearsay, and Apple has not argued that any exceptions
`
`apply. Moreover, there is no proof in the record (from, for example, a librarian)
`
`establishing that any of these references are printed publications. They should thus
`
`be given zero weight, as should the arguments upon which Apple relies on them in
`
`support.
`
`
`
`4
`
`

`

`IPR2021-00970
`
`U.S. Patent No. 9,572,499
`PATENT OWNER’S SUR-REPLY
`
`Apple’s other exhibits are equally problematic. Exhibits 1060-1064 are new
`
`
`
`patents and patent publications—technical evidence with no substantiating expert
`
`testimony. See, e.g., Gemtron Corp. v. Saint-Gobain Corp., 572 F.3d 1371, 1380
`
`(Fed. Cir. 2009) (“[U]nsworn attorney argument ... is not evidence and cannot
`
`rebut ... other admitted evidence....”); Invitrogen Corp. v. Clontech Labs., Inc., 429
`
`F.3d 1052, 1068 (Fed. Cir. 2005).
`
`III. THE PRIOR ART DOES NOT RENDER ARRHYTHMIA
`DETECTION OBVIOUS
`The crux of Apple’s Reply is that the prior art teaches arrhythmia detection.
`
`Apple, however, incorrectly minimizes AliveCor’s arguments and raises various
`
`strawmen, all of which fail.
`
`Apple primarily relies on Shmueli’s disclosure of irregular heart conditions.
`
`Since arrhythmia is an irregular heart condition, Apple argues, Shmueli necessarily
`
`discloses arrhythmia. Reply, 3. But it is well-settled that the disclosure of a broad
`
`genus does not disclose “every species that is a member of that genus.” Atofina v.
`
`Great Lakes Chem. Corp., 441 F.3d 991, 999 (Fed. Cir. 2006). Shmueli does not
`
`give any direction to a POSITA that would lead to the conclusion that arrhythmia is
`
`one of the irregular heart conditions it detects. Resp., 42. Instead, Shmueli is
`
`directed to heart attacks. See, e.g., Ex. 2016, ¶ 64. AliveCor submitted post-
`
`institution evidence with its Response that the Board did not have the benefit of at
`
`
`
`5
`
`

`

`IPR2021-00970
`
`U.S. Patent No. 9,572,499
`PATENT OWNER’S SUR-REPLY
`
`the pre-institution stage, plainly establishing that the disclosure of generic “irregular
`
`
`
`heart conditions” does not disclose any and all conditions that may be considered
`
`irregular. Id. Apple has failed to address that evidence or to address the proper legal
`
`framework to determine when a genus may render a species obvious.
`
`A. AliveCor’s Expert Did Not Agree That The Prior Art Teaches Or
`Suggests Arrhythmia Detection
`Dr. Efimov’s testimony does not “reinforce the Petition” as Apple suggests.
`
`Reply, 3. Instead, Dr. Efimov’s testimony merely reinforced the undisputed
`
`irregular heart condition/arrhythmia genus/species relationship. See Ex. 1069, 28:3-
`
`24 (testifying that arrhythmia is a “subcategory” of irregular heart conditions). See
`
`Atofina, 441 F.3d at 999.
`
`Apple also claims Dr. Efimov incorrectly testified that “sinus tachycardia” is
`
`“an indication of a neurological condition” because that testimony “contradicts
`
`undisputed disclosures in the ’499 patent.” Reply, 3 n.1 (citing Ex. 1069, 58:1-2).
`
`Apple is wrong. Dr. Efimov was plainly referring to Osorio’s disclosure of
`
`“respiratory sinus arrhythmia” (Ex. 1005, [0043]), which he explained in his
`
`declaration is not a cardiac arrhythmia, but is indicative of a neurological condition.
`
`See Ex. 2016, ¶ 9. Respiratory sinus arrhythmia is not disclosed or discussed in
`
`the ’499 patent. See Ex. 1001, Table 2 (discussing tachyarrhythmias such as atrial
`
`fibrillation).
`
`
`
`6
`
`

`

`
`
`IPR2021-00970
`
`U.S. Patent No. 9,572,499
`PATENT OWNER’S SUR-REPLY
`
`B. Dr. Chaitman’s Shmueli Admissions Were Not Limited To
`Embodiments
`Apple cites to AliveCor’s introduction and claims that AliveCor cites no
`
`evidence for its contentions that “Dr. Chaitman admitted that detection in Shmueli
`
`is performed based on the SpO2 measurement and the SpO2 measurement only.”
`
`Reply, 4. This argument is facially incorrect—AliveCor provided specific citation
`
`later in its Response. Response, 51 (citing Ex. 2017, 87:19-88:13). Dr. Chaitman’s
`
`testimony was unambiguous—Shmueli teaches “the detection of the irregular heart
`
`conditions is using [] SpO2 measurements,” not ECG. Id.
`
`Apple then raises a strawman: “[t]o the extent that AliveCor’s position is that
`
`Shmueli’s detection of irregular heart conditions is based only on blood gas
`
`composition measurements, Shmueli’s disclosure and Dr. Efimov’s testimony refute
`
`it.” Reply, 5. That is not AliveCor’s contention. Rather, AliveCor’s position is that
`
`Shmueli’s disclosure that PPG measurements include data, such as pulse rate, that
`
`can be used to detect arrhythmia, is insufficient. Nothing in Shmueli indicates to a
`
`POSITA that its pulse rate data is used to detect arrhythmias. Ex. 2016, ¶ 49.
`
`Moreover, Shmueli mentions pulse rate exactly twice—everywhere else it refers
`
`instead to SpO2 measurements. See id. As AliveCor established in its Response,
`
`based on these scant disclosures, a POSITA would have understood Shmueli’s
`
`
`
`7
`
`

`

`IPR2021-00970
`
`U.S. Patent No. 9,572,499
`PATENT OWNER’S SUR-REPLY
`
`irregular heart conditions to refer to those traditionally detected using SpO2
`
`
`
`measurements, such as heart attacks. See id.; see also Response, 31.
`
`C. Apple Mischaracterizes AliveCor’s Arguments And Shmueli’s
`Disclosures
`Apple argues that Shmueli’s disclosure of an “irregular heart condition” is
`
`“intended to include all such new technologies a priori.” Ex. 1004, 16:3-5. Apple
`
`claims AliveCor “tries to avoid this disclosure” which it calls an “expansive
`
`definition.” Reply, 6 (citing Resp. 53 n.3). Whether this statement is definitional or
`
`“expansive” or not is beside the point. First, this statement is merely aspirational,
`
`and without more (i.e., supporting expert testimony or objective evidence), cannot
`
`be used to conclude that the claims are obvious. Raytheon Techs. Corp. v. Gen. Elec.
`
`Co., 993 F.3d 1374, 1376 (Fed. Cir. 2021) (in reversing a PTAB determination of
`
`obviousness, finding “futuristic” or aspirational statements in the prior art without
`
`supporting evidence could not support a finding obviousness).
`
`Indeed, Apple takes an overbroad and unsubstantiated read of the Shmueli
`
`disclosure. A prior art’s disclosure must enable the claimed invention to render a
`
`patent’s claims obvious. Id. at 1377 (“[I]f an obviousness case is based on a non-
`
`self-enabled reference, and no other prior art reference or evidence would have
`
`enabled a skilled artisan to make the claimed invention, then the invention cannot be
`
`said to have been obvious.”). Shmueli’s disclosure of the broad genus of detecting
`
`
`
`8
`
`

`

`IPR2021-00970
`
`U.S. Patent No. 9,572,499
`PATENT OWNER’S SUR-REPLY
`
`“irregular heart conditions” using PPG does not enable the narrow species of
`
`
`
`arrhythmia detection, which has unique technical challenges. See Ex. 2016, ¶ 64;
`
`see also id., ¶¶ 13-18.
`
`Moreover, Dr. Efimov’s testimony that a POSITA would never have
`
`understood, from Shmueli’s generic and overly inclusive statement of “all such new
`
`technologies” that Shmueli’s device could specifically be used for arrhythmia
`
`detection, is unrebutted. Ex. 2016, ¶ 64; see also id. n.10 (addressing the “new
`
`technologies” disclosure and testifying “[i]n my opinion a POSITA would not read
`
`Shmueli this way.”). Here, Apple argues against another strawman, claiming
`
`AliveCor interprets Shmueli to exclude arrhythmia. Reply, 9. Shmueli does not
`
`expressly exclude arrhythmia, but it also doesn’t include it for the reasons just
`
`discussed.
`
`D.
`Furthermore, Apple’s Positions Are Driven By Hindsight
`Apple argues that because Shmueli teaches “measuring the ECG signal
`
`associated with an intermittent irregular heart-related event,” and because the ’499
`
`patent explains that arrhythmias can be intermittent and infrequent, that Shmueli
`
`teaches arrhythmia. Reply, 7. In so doing, Apple uses the ‘499 patent disclosures as
`
`a roadmap. This is of course impermissible. TQ Delta, LLC v. CISCO Sys., Inc.,
`
`942 F.3d 1352, 1361 (Fed. Cir. 2019). Apple’s reasoning—that because the ’499
`
`
`
`9
`
`

`

`IPR2021-00970
`
`U.S. Patent No. 9,572,499
`PATENT OWNER’S SUR-REPLY
`
`patent and Shmueli use similar language (“intermittent,” “continuously”), the ’499
`
`
`
`patent and Shmueli must both relate to arrhythmia detection—is plainly hindsight
`
`driven. See id. It is a transparent attempt to fit the square peg of Shmueli’s
`
`disclosures into the round hole of the ’499 patent’s claims and should be rejected
`
`Apple does not address AliveCor’s hindsight allegations in reply, instead
`
`claiming AliveCor “simply rehashed generalized arguments from its Preliminary
`
`Response.” Reply, 14. This is not the case as is clear from a review of the post-
`
`institution record. As AliveCor has explained, Apple and its expert Dr. Chaitman
`
`have incorrectly assumed, with no basis in the reference, that Shmueli’s “irregular
`
`heart condition” refers to (or “encompasses”) arrhythmia. This is definitionally
`
`hindsight—Apple is reading disclosures into the prior art in order to map that art to
`
`the claims. See Grain Processing Corp. v. Am. Maize-Prods. Co., 840 F.2d 902, 907
`
`(1988).
`
`E.
`
`Shmueli’s Disclosures Teach A POSITA That Its System Is
`Directed To Heart Attacks, Not Arrhythmias
`In Section II.B.1 of its Reply, Apple again inverts the inquiry. The question
`
`is not whether the “irregular heart conditions” in Shmueli exclude arrhythmia—
`
`Apple’s strawman—but instead whether Shmueli’s disclosures teach arrhythmia
`
`detection to a POSITA. The Board’s decision in the ID that “the present record”
`
`evidence “established” that “irregular heart conditions” “encompass[]” arrhythmia
`
`
`
`10
`
`

`

`IPR2021-00970
`
`U.S. Patent No. 9,572,499
`PATENT OWNER’S SUR-REPLY
`
`is no longer pertinent—AliveCor submitted new evidence, in the form of expert
`
`
`
`testimony, rebutting Apple’s allegations. See, e.g., Ex. 2016, ¶ 64. AliveCor
`
`submits that the record evidence demonstrates Shmueli does not teach or suggest
`
`detecting arrhythmias. Instead, it teaches a POSITA detecting heart attacks.
`
`The Reply then purports to address AliveCor’s “three key arguments,”
`
`claiming each “lacks evidentiary support.” Reply, 10. Apple mischaracterizes
`
`AliveCor’s arguments and the record evidence and, furthermore, is wrong on the
`
`law.
`
`1.
`
`There Is No Record Evidence To Contradict Dr. Efimov
`That “Irregular Heart Condition” Refers To Heart Attacks
`The evidence establishes that Shmueli’s broad, non-specific disclosure of a
`
`broad category of “irregular heart conditions” does not convey arrhythmia detection
`
`to a POSITA. See, e.g., Ex. 2016, ¶¶ 64-68. Instead, a POSITA reading Shmueli
`
`would understand that its system is directed at detecting the types of diseases
`
`traditionally detected using SpO2 measurements—heart attacks. Id. Apple’s expert
`
`Dr. Chaitman has no response and, indeed, Apple has no expert testimony
`
`contradicting AliveCor’s expert testimony that a POSITA would understand
`
`Shmueli is directed to detecting heart attacks even if Dr. Stultz’s impermissible
`
`testimony is considered. See Reply, 10.
`
`
`
`11
`
`

`

`IPR2021-00970
`
`U.S. Patent No. 9,572,499
`PATENT OWNER’S SUR-REPLY
`
`In reply, Apple argues “irregular heart condition” being a genus and
`
`
`
`arrhythmia a species supports Apple’s argument that Shmueli discloses arrhythmia
`
`detection. Reply, 10-11. Apple gets this backwards. It is well established that the
`
`disclosure of a genus in the prior art is not necessarily a disclosure of every species
`
`that is a member of that genus. See, e.g., In re Baird, 16 F.3d 380, 382 (Fed. Cir.
`
`1994). Thus, “[t]here may be many species encompassed within a genus that are
`
`not disclosed by a mere disclosure of the genus.” Ex Parte Roorda, No. 2008-4699,
`
`2009 WL 851420, at *3 (B.P.A.I. Mar. 27, 2009) (emphasis added). Here, as Apple
`
`highlights in its Reply, AliveCor established “there are numerous other irregular
`
`heart conditions,” including heart attacks. Ex. 2016, ¶ 64. Shmueli is not directed
`
`to arrhythmias and its disclosure of a broad genus does not change that fact.
`
`2.
`
`Apple Mischaracterizes AliveCor’s SpO2 Detection
`Arguments And Argues Against Yet Another Strawman
`In Section II.B.1.b, Apple raises yet another strawman, alleging that AliveCor
`
`“assumes no possible configuration for using SpO2 (PPG oximetry) measurements
`
`for arrhythmia detection exists.” Reply, 11. This wholly mischaracterizes
`
`AliveCor’s position, Dr. Efimov’s testimony, and the undisputed evidence. Apple
`
`is correct that “PPG was a ‘suboptimal’ tool for measuring arrhythmias”—this was
`
`conceded by Apple’s ITC expert and is beyond dispute. Ex. 2018 at 62:9-21.
`
`Indeed, “suboptimal replacement” are Apple’s expert’s words. Id. It is indisputable
`
`
`
`12
`
`

`

`IPR2021-00970
`
`U.S. Patent No. 9,572,499
`PATENT OWNER’S SUR-REPLY
`
`that the ECG P-waves are not present in a PPG measurement. Ex. 2016, ¶ 44. But
`
`
`
`neither of these facts address the relevant issue—what would a POSITA have
`
`understood Shmueli’s SpO2 measurement to be used for? The answer is detection
`
`of myocardial infarctions, also known as heart attacks. Ex. 2016, ¶¶ 29-30. Apple
`
`offers no rebuttal to this point.
`
`More problematic, Apple once again resorts to modifying its prima facie
`
`obviousness arguments using new references in a failed attempt to rebut AliveCor’s
`
`arguments. Here, Apple expressly shifts its focus to a secondary reference called
`
`Amano (Ex. 1020) which it argues “discloses an arrhythmia detecting apparatus with
`
`a ‘pulse wave detecting means’” with a “wrist-worn device.” Reply, 12-13. Apple
`
`then claims, with nothing but attorney argument, that “Amano illustrates the
`
`existence of wrist-worn devices configured to use SpO2/PPG data to detect
`
`arrhythmias before the relevant dates.” Reply, 13-14. As already discussed above,
`
`this reference should be struck and Apple’s arguments based on this reference
`
`ignored. But even if Apple’s arguments are considered, they are simply irrelevant.
`
`3.
`
`Shmueli’s CO2 Detection Disclosures Plainly indicate to a
`POSITA that it is Directed To Heart Attacks
`Apple last argues that , that Shmueli’s CO2 measurements are “limited to
`
`specific embodiments” and therefore should be disregarded. Reply, 14. This misses
`
`the point. The point, as explained in AliveCor’s Response, is that CO2
`
`
`
`13
`
`

`

`IPR2021-00970
`
`U.S. Patent No. 9,572,499
`PATENT OWNER’S SUR-REPLY
`
`measurements are understood by POSITAs to be used for heart attack detection and
`
`
`
`not arrhythmia detection. Ex. 2016, ¶ 67. Whether CO2 is an alternative to SpO2 in
`
`Shmueli is irrelevant, because it provides additional context to Shmueli’s
`
`disclosures, further establishing that Shmueli is directed to detecting heart attacks.
`
`Id.
`
`F. Osorio Is Directed To Neurological Conditions, Not Arrhythmia
`Apple raises three flawed arguments in connection with Osorio: (1) Osorio’s
`
`embodiments are non-limiting; (2) claim differentiation merits a finding that
`
`pathological conditions include arrhythmia; and (3) Osorio describes arrhythmia
`
`detection. Reply, 14-16. Apple is incorrect on all counts.
`
`First, Apple is correct that Osorio uses illustrative evidence to describe its
`
`embodiments. But this, again, misses the mark. Every example described in Osorio
`
`as a “pathological condition” is a neurological condition. Ex. 2016, ¶¶ 69-72. And
`
`in the one instance where Osorio teaches “cardiac arrhythmias,” it distinguishes
`
`those arrhythmias from the pathological state, which it describes as a seizure. Ex.
`
`1005, [0071] (“In one embodiment, the severity 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,
`
`
`
`14
`
`

`

`IPR2021-00970
`
`U.S. Patent No. 9,572,499
`PATENT OWNER’S SUR-REPLY
`
`brain electrical activity, the emergence of one or more cardiac arrhythmias, etc.).”)
`
`
`
`(emphasis added); Ex. 2016, ¶ 71.
`
`Furthermore, as with Shmueli, Apple reads too much into Osorio’s generic
`
`disclosure. Osorio’s “pathological conditions” do not enable arrhythmia detection.
`
`See Raytheon, 993 F.3d at 1377 (“[I]f an obviousness case is based on a non-self-
`
`enabled reference, and no other prior art reference or evidence would have enabled
`
`a skilled artisan to make the claimed invention, then the invention cannot be said to
`
`have been obvious.”). Osorio’s generic disclosure of “any pathological state” does
`
`not enable arrhythmia detection, rather Osorio only addresses neurological, not
`
`cardiac, conditions. See Ex. 2016, ¶¶ 69-74.
`
`Second, Apple is wrong on the application of claim differentiation. While it
`
`is true that “the presence of a dependent claim that adds a particular limitation gives
`
`rise to a presumption that the limitation in question is not present in the independent
`
`claim,” Phillips v. AWH Corp., 415 F.3d 1303, 1314-15 (Fed. Cir. 2005), Apple is
`
`wrong to suggest that claim differentiation changes the outcome here. Osorio claim
`
`7 recites that the “pathological body state is an epileptic event.” This does not
`
`preclude AliveCor’s reading of Osorio—that the pathological states described in
`
`Osorio are neurological, and not cardiological—since the broader independent claim
`
`would include other neurological conditions. See Resp., 47-48.
`
`
`
`15
`
`

`

`IPR2021-00970
`
`U.S. Patent No. 9,572,499
`PATENT OWNER’S SUR-REPLY
`
`Third, Apple is incorrect that Osorio “describes arrhythmia detection”—it
`
`
`
`does not. AliveCor’s expert Dr. Efimov addressed each instance of Osorio’s
`
`references to tachycardia and arrhythmia and explained how in every instance
`
`Osorio described the pathological state as an epileptic event. Ex. 2016, ¶¶ 72-74.
`
`Apple makes no effort to address Dr. Efimov’s testimony regarding Osorio’s
`
`disclosures as being consistent with and teaching a POSITA that its system relates
`
`to neurological conditions such as epilepsy and, as such, his opinions are unrebutted.
`
`IV. A POSITA WOULD NOT HAVE BEEN MOTIVATED TO COMBINE
`SHMUELI AND OSORIO
`Apple erroneously asserts that AliveCor “does not dispute any of the
`
`motivations to combine Shmueli and Osorio identified in the Petition.” Reply, 23.
`
`This is untrue. In the Petition, Apple asserted that Osorio “describes the benefits …
`
`of using activity level to detect an irregular heart condition.” Pet., 20. Apple asserts
`
`that the combination would incorporate pathological events “e.g., arrhythmia” from
`
`Osorio. Id. All of the Petition’s alleged motivations are based on the premise that
`
`the two references disclose arrhythmias. See id., 20-31. Thus, by arguing that
`
`neither Shmueli or Osorio disclose arrhythmia detection, AliveCor undercuts the
`
`foundation for all of Apple’s motivations to combine the two references.
`
`Despite Apple’s protestations, both Adidas and Nichia are on point. The
`
`Shmueli/Osorio combination would “alter[ ] the principles of operation” of both
`
`
`
`16
`
`

`

`IPR2021-00970
`
`U.S. Patent No. 9,572,499
`PATENT OWNER’S SUR-REPLY
`
`references, neither of which are directed to arrhythmia detection, so that neither
`
`
`
`would work for its intended purpose, just as in Adidas. See Adidas AG v. Nike Inc.,
`
`963 F.3d 1355, 1359 (Fed. Cir. 2020). The two references are fundamentally
`
`different and a POSITA would not combine them. Id. And in Nichia, the Court
`
`found no motivation to combine where the references disclosed “different structures,
`
`resolve[d] dissimilar problems, and propose[d] dissimilar solutions.” Nichia Corp.
`
`v. Everlight Ams., Inc., 855 F.3d 1328, 1340 (Fed. Cir. 2017). Shmueli and Osorio
`
`resolve dissimilar problems—heart attacks in Shmueli and neurological conditions
`
`in Osorio—using different solutions—SpO2 measurement in Shmueli and activity
`
`level monitoring in Osorio. Ex. 2016, ¶¶ 46-57, 76-77.
`
`V. NONE OF THE CITED ART RENDERS OBVIOUS MACHINE
`LEARNING ALONE OR IN COMBINATION
`Apple first argues that the machine learning dependent claims are “nothing
`
`more than generic functional language that adds no inventive concept.” Reply, 18.
`
`To the extent Apple is raising a claim construction argument, it has failed to
`
`provide the proper legal analysis. Whether a term is functional depends on a rigorous
`
`analysis, none of which Apple provides here. See, e.g., Nevro Corp. v. Boston
`
`Scientific Corp., 955 F.3d 35, 39 (Fed. Cir. 2020). Apple’s “functional language”
`
`attack on the machine learning claims should therefore be rejected. Indeed, these
`
`terms are plainly not functional—the claims recite “determining a presence of said
`
`
`
`17
`
`

`

`IPR2021-00970
`
`U.S. Patent No. 9,572,499
`PATENT OWNER’S SUR-REPLY
`
`arrhythmia using a machine learning algorithm.” The claims themselves thus
`
`
`
`plainly recite structure, which plainly establish that the limitations are non-
`
`functional.
`
`Setting aside Apple’s new claim construction argument, it is clear that the
`
`prior art does not disclose the required limitations. Dr. Efimov has offered testimony
`
`that the types of machine learning required by the ‘499 patent were not known to be
`
`applicable to arrhythmia detection as the claims require. See Ex. 1069, 173:19-
`
`175:1. Apple’s only response is cite several paragraphs of Dr. Chaitman’s Petition
`
`testimony regarding motivation to combine and a handful of references allegedly
`
`establishing that the claimed machine learning was used in other contexts, and which
`
`are not part of the challenged grounds. See Reply, 18 (citing Ex. 1003, ¶¶192-199;
`
`Exs. 1012, 1038-1041, 1071, 1073).
`
`But mere knowledge of a technique is not a motivation to modify an existing
`
`solution to use that technique. This is tantamount to arguing that a POSITA could
`
`have combined the references, not that they would have, and obviousness requires
`
`more. Polaris Indus., Inc. v. Arctic Cat, Inc., 882 F.3d 1056, 1068 (Fed. Cir. 2018)
`
`(criticizing the Board’s analysis as focusing on “what a skilled artisan would have
`
`been able to do, rather than what a skilled artisan would have been motivated to do
`
`
`
`18
`
`

`

`IPR2021-00970
`
`U.S. Patent No. 9,572,499
`PATENT OWNER’S SUR-REPLY
`
`at the time of the invention”); Belden Inc. v. Berk-Tek LLC, 805 F.3d 1064, 1073
`
`
`
`(Fed. Cir. 2015).
`
`Moreover, Apple dismisses out of hand Dr. Efimov’s testimony that there is
`
`tremendous industry skepticism in the field regarding the use of machine learning.
`
`Ex. 2016, ¶ 85. Apple introduces new evidence to rebut Dr. Efimov, but Apple
`
`offers no POSITA testimony in support, instead relying exclusively on attorney
`
`argument. See Reply, 19. This attorney argument is not a rebuttal to Dr. Efimov’s
`
`substantiated expert testimony. Gemtron, 572 F.3d at 1380.
`
`Finally, Apple seeks to focus the board on the “benefits” provided by machine
`
`learning. Reply, 19-20. Again, however, these “benefits,” even if substantiated
`
`(they are nothing but attorney argument) merely amount to argument the
`
`combination “could have” been made. There is no evidence in the record that a
`
`POSITA would have had a reason to make this combination. See TQ Delta, 942
`
`F.3d at 1361.
`
`Apple then makes three arguments specific to its machine learning grounds:
`
`(1) the claims do not limit machine learning to PPG data; (2) a POSITA would have
`
`found it