Zepp Health Corp v. Worcester Polytechnic Institute

1. Case Identification

• Case #: IPR2025-00522
• Patent #: 9,713,428
• Filed: February 4, 2025
• Petitioner(s): Zepp Health Corp
• Patent Owner(s): Worcester Polytechnic Institute; THE RESEARCH FOUNDATION FOR THE STATE UNIVERSITY OF NEW YORK
• Challenged Claims: 1-11, 15-27, 29-30, 32-37, 39-40, 42-46

2. Patent Overview

• Title: Physiological Parameter Monitoring
• Brief Description: The ’428 patent discloses systems and methods for analyzing physiological signals, such as photoplethysmographic (PPG) and electrocardiogram (ECG) signals, from mobile devices. The invention applies statistical methods to preprocess signals and detect motion artifacts or atrial fibrillation (AF) by comparing results, such as Shannon Entropy or a time-varying coherence function, against a threshold.

3. Grounds for Unpatentability

Ground 1: Claims 1, 3-11, 15, 21, 23-27, 29-30, 37, 39-40 are obvious over Asada in view of Chon-570.

• Prior Art Relied Upon: Asada (a 2003 journal article on wearable PPG biosensors) and Chon-570 (WO Publication # 2009/018570).
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner argued that Asada taught a wearable biosensor (WBS) system, such as a ring sensor with a processor, that monitors physiological parameters (e.g., heart rate, oxygen saturation) and transmits data to a handheld device like a PDA or cellphone. Asada recognized the problem of motion artifacts and disclosed methods to reduce them. Petitioner asserted Chon-570 taught analyzing physiological signals (e.g., ECG data) using statistical methods, including the specific Shannon Entropy formula recited in independent claim 1, to detect outliers and conditions like AF. The combination allegedly rendered the core limitations of claim 1 obvious: providing a physiological signal to a handheld device (Asada), analyzing it to obtain physiological parameters (Asada, Chon-570), and detecting motion artifacts using Shannon Entropy compared against a threshold (Chon-570).
  • Motivation to Combine: A POSITA would combine these references because they are in the same field of real-time physiological monitoring. It would have been an obvious application of a known statistical technique (Shannon Entropy from Chon-570) to a known WBS system (Asada) to improve its capabilities for detecting motion artifacts, a problem Asada explicitly sought to solve. This combination would yield a predictable result.
  • Expectation of Success: The algorithm in Chon-570 was platform-agnostic and could readily be implemented on the processing hardware disclosed in Asada’s WBS system, ensuring a high expectation of success.

Ground 2: Claims 1-11, 15, 21-27, 29-30, 37, 39-40 are obvious over Asada and Chon-570 in view of Delorme.

• Prior Art Relied Upon: Asada, Chon-570, and Delorme (a 2003 conference proceeding on artifact rejection in EEG data).
• Core Argument for this Ground:
  • Prior Art Mapping: This ground built upon Ground 1, adding Delorme to further support the obviousness of detecting motion artifacts. Petitioner argued Delorme explicitly taught using higher-order statistics, including both Shannon Entropy and Kurtosis, to detect and reject motion artifacts from physiological signals. Delorme disclosed using rejection thresholds based on these statistical measures to decide whether to retain data segments. The addition of Delorme was argued to render claim 2, which recites using kurtosis as an indicator of volatility, obvious.
  • Motivation to Combine: A POSITA seeking to improve Asada's motion artifact detection would have been motivated to incorporate the higher-order statistical methods taught by Delorme. Furthermore, a POSITA would have been motivated to substitute the general entropy calculation in Delorme with the more specific and rigorous Shannon Entropy formula from Chon-570 to better detect irregularities in physiological data.
  • Expectation of Success: Combining these known signal processing techniques for their intended purposes—artifact detection and data analysis—on a capable hardware platform like Asada's would have been straightforward for a POSITA.

Ground 3: Claims 16-20, 32-36, 42-46 are obvious over Asada and Chon-570 in view of Chon-2008.

• Prior Art Relied Upon: Asada, Chon-570, and Chon-2008 (a 2008 journal article on analyzing nonstationarity in physiological data).
• Core Argument for this Ground:
  • Prior Art Mapping: This ground targeted the "TVCF Claims," which recite detecting AF by obtaining a time-varying coherence function (TVCF). Petitioner asserted Asada and Chon-570 taught monitoring for arrhythmia and AF. Chon-2008 taught the specific claimed method for obtaining a TVCF: multiplying two time-varying transfer functions (TVTFs), where the second TVTF is derived by reversing the input and output signals of the first TVTF. Chon-2008 also taught comparing the resulting TVCF to a predetermined threshold to analyze nonstationarity.
  • Motivation to Combine: A POSITA would combine these references to improve AF detection in the Asada/Chon-570 system. Since AF introduces non-stationary elements into physiological signals, a POSITA would have been motivated to apply the known statistical method for analyzing nonstationarity from Chon-2008 (deriving a TVCF) to the problem of detecting AF. This would be an obvious application of a known tool to a known problem.
  • Expectation of Success: Applying the mathematical techniques from Chon-2008 to the physiological data acquired by the Asada system was a predictable implementation of signal processing principles with a high expectation of success.

4. Arguments Regarding Discretionary Denial

• §314(a) / Fintiv: Petitioner argued against discretionary denial under Fintiv, asserting that the parallel district court litigation was in its early stages with minimal investment from the parties and no substantive orders issued. Petitioner noted that the scheduled trial date is likely to be delayed due to the judge's heavy caseload. Crucially, Petitioner filed a stipulation agreeing not to pursue in the district court any grounds raised or that reasonably could have been raised in the IPR, which Petitioner argued weighs decisively against denial per USPTO guidance.
• §325(d): Petitioner argued against denial under §325(d), stating that the examiner erred in allowing the claims. The petition asserted that the primary prior art references (Asada, Chon-570, Chon-2008) were never considered during prosecution. Petitioner contended that these new references teach the key elements—including the specific Shannon Entropy formula and the use of a PPG sensor—that the patentee added to the claims to overcome prior rejections. Therefore, the examiner did not have the benefit of the strongest prior art combinations.

5. Relief Requested

• Petitioner requested the institution of an inter partes review and the cancellation of claims 1-11, 15-27, 29-30, 32-37, 39-40, and 42-46 of the ’428 patent as unpatentable.