Walk the independent claim. Fitbit's grant US10973422B2, "Photoplethysmography-based pulse wave analysis using a wearable device" (issued April 13, 2021; inventors Alexandros Pantelopoulos and Andrew Axley), is a granted patent. Its CPC anchor A61B 5/02438 is the photoplethysmography (PPG) class — optical pulse sensing — surrounded by cardiovascular-measurement classes that signal the claim reaches past heart rate into vascular health.

Claim 1 builds a "biometric monitoring device for measuring arterial stiffness" from four parts: a wearable fixing structure that attaches to the user during activity, an inertial sensor that measures the device's movement, a PPG sensor, and one or more processors. The processors are where the invention lives, and the claim recites a specific gated, motion-aware acquisition strategy. The device runs in a first mode when the user is performing one activity — and in that first mode the processors either do not obtain PPG data at all, or obtain it at a lower sampling rate. They obtain inertial data while the user performs a second activity, and from that inertial data they determine two things: that the device has experienced movement below a movement threshold for a period of time, and that the inertial data matches an "orientation profile." Only when both conditions are met does the device switch to a second mode to obtain PPG data. It then filters the PPG signal using information from the inertial data, and determines one or more "morphological features of a pulse waveform" from the filtered signal — explicitly "without using electrode-based heart sensor data" — where those morphological features are "related to arterial stiffness."

“Disclosed are devices and methods for non-invasively measuring arterial stiffness using pulse wave analysis of photoplethysmogram data.”— U.S. Patent No. 10,973,422 source

Two limitations carry the weight. The first is the inertial gating: the device opportunistically waits for a quiet, correctly-oriented moment — low motion below a threshold, matching an orientation profile such as a wrist held still — before it bothers to capture high-quality PPG. That is what makes a motion-corrupted wrist signal usable for waveform shape analysis, which is far more motion-sensitive than simple beat counting. The second is the analysis target itself: the claim does not stop at heart rate; it extracts the morphology of the pulse wave — its shape, rise, and contour — and explicitly does so optically, without an ECG electrode. Arterial stiffness is inferred from how the pressure wave's shape changes, so the claim is squarely about reading vascular properties from the optical waveform, not about counting pulses.

The dependent claims sharpen both ideas. Claim 2 enumerates the activity types — resistance training, aerobic exercise, endurance training, sitting, working, sleeping — that drive the mode logic. Claim 3 fixes the first mode at a lower sampling rate than the second. Claim 4 adds temperature, strain, or pressure sensors and performs "wave normalization" on the pulse waveform using that data — correcting the waveform for confounders before features are read. Claim 5 uses the inertial data to reject motion artifact directly. Claims 6 through 9 govern how often the device triggers a capture: claim 7 makes the number of pulse waveforms collected depend on the user's heart rate, claim 8 makes it depend on respiration rate, and claim 9 schedules captures by activity type or location. Claim 10 closes the loop by analyzing the waveform to estimate arterial stiffness. The two further independent claims (11 and 12) recast the device around triggering a number of pulse waveforms that depends on heart rate (11) or respiration rate (12), again filtering with inertial data and extracting stiffness-related morphological features.

So the element doing the work is the inertially-gated capture of an optical pulse waveform followed by morphological feature extraction tied to arterial stiffness — not merely counting beats, but waiting for a clean moment, cleaning the signal with motion data, and reading the wave's shape for a vascular metric. That is the difference between a step counter with a heart-rate readout and a wrist device making inferences about cardiovascular health from an optical sensor alone.

What it reads on is the modern fitness wearable's health-metric stack: the features built on top of optical pulse sensing that go beyond beats-per-minute. As a Fitbit (and now Google) asset, the claim sits inside a portfolio whose premise is turning a wrist optical sensor into a health platform.

Scope discipline: PPG itself is established medical art, and the claim does not own it. What it protects is the recited combination — motion-gated PPG acquisition, inertial filtering, and morphological pulse-wave feature extraction related to arterial stiffness, performed without electrode data. A device that reads PPG only for heart rate, without the inertial gating and waveform-morphology analysis, operates in different territory. A chest-strap system that derives the same metrics from ECG electrodes also falls outside, since claim 1 expressly excludes electrode-based heart-sensor data. The defensible element is the gated, motion-corrected waveform-morphology analysis, not the optical sensing.

Granted status places US10973422B2 among the enforceable assets in the crowded wrist-health space — a thicket that also includes Apple, Samsung, Masimo, and Valencell. For a freedom-to-operate read on any wrist-based cardiovascular metric, the pulse-wave-analysis claims, and specifically the inertial-gating limitation, are the ones to map against. The portfolio signal is consistent with Fitbit's lineage: optical sensing as the spine of a health story, with this grant the analysis layer — the claim that turns a raw, motion-prone optical signal into a defensible vascular-health inference.