DCT

2:23-cv-00353

Research Foundation for State University Of New York v. Xiaomi Corp

Log In

I. Executive Summary and Procedural Information

  • Parties & Counsel:
  • Case Identification: 2:23-cv-00353, E.D. Tex., 09/11/2023
  • Venue Allegations: Plaintiffs allege venue is proper because Defendants are subject to personal jurisdiction in the district, have committed acts of infringement there, maintain sufficient minimum contacts with the district, and are foreign corporations, against whom suit is proper in any judicial district.
  • Core Dispute: Plaintiffs allege that Defendants’ smartwatches and associated software applications infringe seven patents related to algorithms for physiological monitoring, such as detecting cardiac arrhythmias and filtering motion artifacts from biometric sensor data.
  • Technical Context: The technology involves processing physiological data from wearable sensors to provide health insights, a key feature set in the rapidly growing consumer smartwatch market.
  • Key Procedural History: The complaint does not mention any prior litigation, Inter Partes Review (IPR) proceedings, or licensing history related to the patents-in-suit.

Case Timeline

Date Event
2007-08-02 ’326 Patent Priority Date
2011-01-21 ’428 Patent Priority Date
2013-04-09 ’326 Patent Issue Date
2013-05-01 ’576 Patent Priority Date
2014-05-01 ’921 and ’601 Patents Priority Date
2015-01-29 ’362 Patent Priority Date
2015-06-09 ’647 Patent Priority Date
2016-08-09 ’576 Patent Issue Date
2017-07-25 ’428 Patent Issue Date
2018-06-05 ’921 Patent Issue Date
2019-05-07 ’647 Patent Issue Date
2019-05-14 ’601 Patent Issue Date
2020-05-19 ’362 Patent Issue Date
2023-09-11 Complaint Filing Date

II. Technology and Patent(s)-in-Suit Analysis

U.S. Patent No. 8,417,326 - RR Interval Monitoring Method and Blood Pressure Cuff Utilizing Same

  • Issued: April 9, 2013

The Invention Explained

  • Problem Addressed: The complaint notes that prior art methods for detecting Atrial Fibrillation (AF) were limited by the need to obtain large databases of training data (Compl. ¶45). The patent’s background section further discusses the need for ambulatory, real-time AF detection methods (’326 Patent, col. 1:15-20).
  • The Patented Solution: The invention provides a method for real-time AF detection by analyzing the statistical properties of heart rate intervals (known as RR intervals) without requiring training data (Compl. ¶45). The method involves calculating a combination of three specific statistical metrics—Turning Points Ratio (TPR), Root Mean Square of Successive Differences (RMSSD), and Shannon Entropy (SE)—from a series of heartbeats to detect the likelihood of AF (’326 Patent, col. 2:50-61).
  • Technical Importance: This approach enabled more accurate, real-time AF detection using statistical techniques that did not rely on large, pre-existing training datasets (Compl. ¶45).

Key Claims at a Glance

  • The complaint asserts at least independent Claim 1 against all Defendants (Compl. ¶120, ¶281).
  • The essential elements of Claim 1 include:
    • An Atrial Fibrillation (AF) analysis method comprising:
    • obtaining an output that includes a heart beat;
    • deriving a heart beat interval;
    • analyzing a number (N) of heart beat intervals from the output; and
    • detecting a likelihood of AF by: calculating a Turning Points Ratio (TPR) of the N heart beat intervals; calculating a root mean square of successive (RMSSD) heart beat intervals; and calculating Shannon Entropy (SE) of the N heart beat intervals.
  • The complaint reserves the right to assert additional claims (Compl. ¶137, ¶298).

U.S. Patent No. 9,408,576 - Detection and Monitoring of Atrial Fibrillation

  • Issued: August 9, 2016

The Invention Explained

  • Problem Addressed: The complaint states the patent addresses inaccurate AF algorithm methods that could not effectively discriminate between AF and other cardiac events like premature atrial/ventricular contractions (PAC/PVC), which can mimic the random dynamics of AF (Compl. ¶54, ¶72).
  • The Patented Solution: The invention is an enhanced algorithm for accurately detecting and discriminating between various heart rhythms, including normal sinus rhythm (NSR), AF, PVC, and PAC (Compl. ¶51). The method involves constructing a "Poincare plot" of time interval data, identifying patterns corresponding to specific arrhythmias (e.g., bigeminy, trigemini), subtracting that data, and then analyzing the remaining "updated data" with statistical measures like RMSSD and Shannon Entropy to determine if AF is present (’576 Patent, col. 2:28-64).
  • Technical Importance: This provided a more efficient and accurate real-time method for AF detection that could distinguish AF from other arrhythmias that produce similar signal characteristics (Compl. ¶54).

Key Claims at a Glance

  • The complaint asserts at least independent Claim 1 against all Defendants (Compl. ¶140, ¶301).
  • The essential elements of Claim 1 include a computer-implemented method comprising demarcating boundaries in a Poincare plot space, constructing a Poincare plot from a subject's heart rate signal, identifying patterns corresponding to PAC or PVC, obtaining updated data by subtracting those patterns, and then performing a statistical analysis on the updated data to determine if the subject has AF or NSR with PVC/PAC (Compl. ¶144, ¶305).
  • The complaint reserves the right to assert additional claims (Compl. ¶164, ¶325).

Multi-Patent Capsules

  • Patent Identification: U.S. Patent No. 9,713,428, Physiological Parameter Monitoring with a Mobile Communication Device, issued July 25, 2017.

  • Technology Synopsis: The patent is directed to systems and methods for physiological health monitoring using a mobile device that can detect motion artifacts in the signal to ensure the reported results are of acceptable quality (Compl. ¶60, ¶63). The method involves analyzing a physiological signal, calculating an indicator of volatility such as Shannon entropy, and deciding whether to retain the measurements based on whether the indicator exceeds a threshold for noise or motion artifacts (Compl. ¶171, ¶332).

  • Asserted Claims: At least Claim 1 (Compl. ¶167, ¶328).

  • Accused Features: The accused watches are alleged to use algorithms that enable physiological monitoring while detecting motion artifacts to ensure data quality (Compl. ¶175, ¶179).

  • Patent Identification: U.S. Patent No. 9,986,921, Detection and Monitoring of Atrial Fibrillation, issued June 5, 2018.

  • Technology Synopsis: The patent describes a real-time arrhythmia discrimination method that can distinguish between NSR, AF, PACs, and PVCs using pulsatile time series data, for example from a smartphone's camera (Compl. ¶69). It addresses the problem of inaccurate AF detection when many PAC/PVC episodes are present, which can mimic the random dynamics of AF (Compl. ¶72).

  • Asserted Claims: At least Claim 1 (Compl. ¶190, ¶351).

  • Accused Features: The accused watches are alleged to implement a method for discriminating between normal sinus rhythm and various arrhythmias by obtaining and comparing statistical measures like RMSSD, Shannon entropy, and turning point ratio to predetermined thresholds (Compl. ¶194-197, ¶355-358).

  • Patent Identification: U.S. Patent No. 10,278,647, Method and Apparatus for Removing Motion Artifacts from Biomedical Signals, issued May 7, 2019.

  • Technology Synopsis: The patent is directed to a method that employs a time-varying spectral analysis approach to reconstruct a heart-related signal that has been corrupted by motion artifacts (Compl. ¶78). This overcomes limitations in accurately estimating heart rate from signals, such as a PPG signal, during intense physical activity (Compl. ¶81). The method involves analyzing both the heart-related signal and a motion signal from a motion sensor to identify and remove the artifacts (Compl. ¶213, ¶374).

  • Asserted Claims: At least Claim 1 (Compl. ¶209, ¶370).

  • Accused Features: The accused watches allegedly use a time-varying spectral approach with both a heart-rate sensor (biomedical sensor) and a motion sensor to reconstruct a heart-related signal that includes motion artifacts (Compl. ¶214-215, ¶375-376).

  • Patent Identification: U.S. Patent No. 10,285,601, Detection and Monitoring of Atrial Fibrillation, issued May 14, 2019.

  • Technology Synopsis: The patent describes a system for real-time arrhythmia discrimination that can distinguish between NSR, AF, PACs, and PVCs using pulsatile time series data collected from a smartphone's camera (Compl. ¶87). Like the ’921 patent, it addresses the inaccurate detection of AF in the presence of many PAC/PVC episodes (Compl. ¶90).

  • Asserted Claims: At least Claim 1 (Compl. ¶229, ¶390).

  • Accused Features: The accused watches are alleged to be a system with one or more processors configured to discriminate between normal sinus rhythm and other arrhythmias by calculating and comparing statistical metrics (RMSSD, Shannon entropy, turning point ratio) against thresholds (Compl. ¶233, ¶394).

  • Patent Identification: U.S. Patent No. 10,653,362, Motion and Noise Artifact Detection and Reconstruction Algorithms for Photoplethysmogram and Equivalent Signals, issued May 19, 2020.

  • Technology Synopsis: The patent is directed to a pulse oximeter with an embedded motion and noise artifact (MNA) detection algorithm (Compl. ¶96). This technology overcomes the limitation of inaccurate heart rate and oxygen saturation detection during body movements by using a time-varying spectral analysis to extract features unique to clean versus corrupted signal components (Compl. ¶99).

  • Asserted Claims: At least Claim 1 (Compl. ¶252, ¶413).

  • Accused Features: The accused watches are alleged to implement a method for monitoring physiological parameters from a PPG signal by obtaining a time frequency spectrum, calculating a noise quality index from it, and using that index to determine if the signal segment is corrupted by motion artifacts (Compl. ¶256-259, ¶417-420).

III. The Accused Instrumentality

Product Identification

  • The accused instrumentalities are various smartwatch models sold under the Xiaomi and Zepp/Amazfit brands, including the “Mi Smart Band” series, “Redmi Watch” series, “Xiaomi Watch” series, “Amazfit Falcon,” “T-Rex” series, and “GTR”/”GTS” series (collectively, the “Accused Watches”) (Compl. ¶121, ¶282). The infringement allegations also extend to the associated mobile software applications, including “Mi Fitness,” “Zepp Life,” and “Zepp” (Compl. ¶104-105, ¶112-114).

Functionality and Market Context

  • The Accused Watches are wearable devices equipped with multi-channel photoplethysmography (PPG) bio-sensors for physiological monitoring (Compl. p. 29). They are marketed as providing features such as "All-day heart rate monitoring," "Abnormal heart rate alert," blood-oxygen saturation monitoring, and workout analysis (Compl. ¶106, ¶108, ¶115, ¶117). The watches connect to the accused mobile applications, which display and analyze the collected health data (Compl. ¶105, p. 30). The complaint alleges a close strategic and manufacturing partnership between Xiaomi and Zepp, with Zepp manufacturing the Mi Band series for Xiaomi and generating a significant portion of its revenue from such sales (Compl. ¶11-14). An image from Xiaomi's website shows the Xiaomi Smart Band 7 measuring heart rate and displaying different rates for intense workouts, high heart rate, everyday state, and sleep (Compl. p. 29).

IV. Analysis of Infringement Allegations

’326 Patent Infringement Allegations

Claim Element (from Independent Claim 1) Alleged Infringing Functionality Complaint Citation Patent Citation
An Atrial Fibrillation (AF) analysis method comprising: obtaining an output that includes a heart beat; The Accused Watches perform "All-day heart rate monitoring" using multi-channel PPG bio-sensors to obtain heart beat data. ¶126; p. 29 col. 2:50-52
deriving a heart beat interval; The watches' algorithms are alleged to derive heart beat intervals from the sensor output. The complaint provides visuals of heart rate data being shown in detail. ¶127; p. 32 col. 2:53-54
analyzing a number (N) of heart beat intervals from the output; and The watches are alleged to analyze a number of these intervals to provide health statistics and alerts. ¶128; p. 34 col. 2:55-56
detecting a likelihood of AF by: calculating a Turning Points Ratio (TPR) of the N heart beat intervals; calculating a root mean square of successive (RMSSD) heart beat intervals; and calculating Shannon Entropy (SE) of the N heart beat intervals. The complaint alleges on information and belief that the underlying algorithms used by the Accused Watches detect a likelihood of AF by performing these three specific calculations. ¶129; ¶290 col. 2:57-61
  • Identified Points of Contention:
    • Evidentiary Question: The complaint alleges on "information and belief" that the defendants' algorithms perform the three specific mathematical calculations recited in the final limitation of Claim 1 (TPR, RMSSD, and SE) (Compl. ¶129). A central point of contention will be an evidentiary one: what proof exists that the accused products' general "Abnormal heart rate alert" functionality (Compl. p. 29) is achieved by implementing these exact claimed statistical methods, as opposed to different, non-infringing algorithms that produce a similar outcome?

’576 Patent Infringement Allegations

Claim Element (from Independent Claim 1) Alleged Infringing Functionality Complaint Citation Patent Citation
A computer implemented method for discriminating between atrial fibrillation and premature ventricular contractions (PVC) and premature atrial contractions (PACs)... The Accused Watches are alleged to utilize algorithms capable of differentiating various patterns of arrhythmias, including PACs and PVCs, from normal sinus rhythm and AF. ¶145; ¶156 col. 1:17-21
demarcating boundaries in a Poincare plot space, the boundaries being obtained from data from a test set of test subjects... The complaint alleges on information and belief that the accused algorithm demarcates boundaries in a Poincare plot space based on data from a test set of subjects. ¶146 col. 2:28-34
constructing a Poincare plot of time interval data from a subject under test... The complaint alleges on information and belief that the accused algorithm constructs a Poincare plot using time interval data from the user's heart rate signal. ¶148 col. 2:35-39
identifying data in patterns in the Poincare plot, the patterns including patterns corresponding to combinations of at least one of bigeminy, trigemini, and quadragemini indicating one of PAC or PVC; On information and belief, the accused algorithm is alleged to identify specific arrhythmia patterns such as bigeminy within the Poincare plot. ¶149 col. 2:40-44
obtaining updated data by subtracting the data in the patterns... from the time interval data from the subject under test; On information and belief, the accused algorithm is alleged to obtain updated data by subtracting the identified patterns from the initial data set. ¶150 col. 2:45-50
obtaining a root mean squared of successive differences, a Shannon entropy and a turning point ratio for the updated data; On information and belief, the accused algorithm is alleged to perform these specific statistical calculations on the remaining "updated data." ¶151 col. 2:51-54
...comparing [the metrics] to... predetermined threshold[s]; determining... that the subject under test has atrial fibrillation; and determining... that the subject under test has normal sinus rhythm (NSR) with PVC or PAC... On information and belief, the accused algorithm is alleged to compare the calculated metrics against thresholds to determine whether the user has AF or NSR with PAC/PVC. ¶152-154 col. 2:55-3:9
  • Identified Points of Contention:
    • Scope and Technical Questions: The claim recites a highly specific, multi-step method involving the creation and manipulation of a "Poincare plot space." A key question of claim scope and technical operation will be whether the accused products' general heart rhythm analysis functionalities (Compl. ¶156) can be shown to technically practice these specific steps. Does the accused algorithm actually construct and subtract patterns from a "Poincare plot" as defined by the patent, or does it use a different analytical framework to discriminate between arrhythmia types?

V. Key Claim Terms for Construction

  • The Term: "Poincare plot space" (from ’576 Patent, Claim 1)
  • Context and Importance: This term is the technological core of Claim 1 of the ’576 Patent. The infringement theory depends entirely on whether the defendants' accused data processing method can be characterized as operating within a "Poincare plot space" as understood in the patent. Practitioners may focus on this term because it appears to be a specific technical tool, and the case may turn on whether the accused algorithm's general analysis of heart rate variability falls within the patent's more specific definition.
  • Intrinsic Evidence for Interpretation:
    • Evidence for a Broader Interpretation: The claim itself defines it as "a space of time interval between consecutive pulses obtained by sensing variability in heart rate signal" (Compl. ¶144; ’576 Patent, col. 20:3-6). This language could be argued to encompass any graphical or data space that plots one heart rate interval against another.
    • Evidence for a Narrower Interpretation: The patent's detailed description and figures may depict only a very specific type of Poincare plot with particular axes and boundary conditions (e.g., ’576 Patent, Fig. 3, Fig. 4a-4d). A defendant may argue that these specific embodiments limit the term to the particular plots disclosed, and not to any generic space of heart rate intervals.

VI. Other Allegations

  • Indirect Infringement: The complaint alleges that Defendants induce infringement by providing product manuals, marketing materials, and software applications that instruct and encourage end-users to use the accused physiological monitoring features (e.g., Compl. ¶136, ¶163, ¶297). It further alleges that Zepp induces Xiaomi’s infringement through their close manufacturing, design, and operational partnership (Compl. ¶27).
  • Willful Infringement: The complaint alleges willfulness based on Defendants' knowledge of the patents-in-suit and their applicability to the accused products since "at least the filing of this complaint" (e.g., Compl. ¶131, ¶158). This frames the allegation as post-suit willfulness.

VII. Analyst’s Conclusion: Key Questions for the Case

  • A key evidentiary question will be one of algorithmic identity: does discovery show that the defendants' software for detecting heart rate irregularities performs the specific mathematical steps recited in the asserted claims (e.g., calculating TPR, RMSSD, and SE for the ’326 patent, or demarcating and subtracting patterns in a Poincare plot space for the ’576 patent), or does it use fundamentally different, non-infringing algorithms to achieve a similar commercial feature?
  • A central issue will be one of claim scope: can the specific, multi-step analytical methods described in the patents, which are rooted in distinct mathematical and signal processing techniques, be construed broadly enough to cover the general-purpose "all-day heart rate monitoring" and "abnormal heart rate alert" functionalities marketed in the accused products?