PulseSensor Research · knowledge-base front door
A connected guide to photoplethysmography science, devices, patents, products, reproducible experiments, and open questions—for people who want to understand PPG well enough to make something real.
Start with your question
The same PPG record can be a physiological signal, an engineering problem, a product implementation, a patent disclosure, or an experiment. Start with the job, then check the evidence type.
Build a defensible foundation in physiology, measurement, algorithms, quality, and validation.
Learn the science → 2Trace light, placement, chips, mechanics, software, and clinical or wearable implementations.
Explore implementations → 3Read technical disclosures with family, product, science, status-date, and legal-uncertainty context.
Understand invention → 4Choose a bounded question, record raw data and references, report failures, and make the work repeatable.
Build and validate →Four knowledge-base doors
Each door exposes existing research collections. Nothing here replaces the annotated bibliography, Read / See / Try lanes, Innovation Graph, or PPG IP Atlas.
Understand what the optical waveform can support, where it fails, and how a result was validated.
Follow the full instrument: tissue site, geometry, wavelengths, analog front end, contact, sampling, algorithms, and packaging.
Use patents for technical and commercial landscape literacy with dated status screening and explicit uncertainty.
Turn an idea into a safe, inspectable protocol and a result someone else can challenge or reproduce.
Cross-cutting index
These tracks are a shared navigation vocabulary. The current index uses curated routes; it does not imply that every legacy Page has complete database tags.
What information are you trying to estimate or inspect?
Where and how does light enter and return from tissue?
Which layer shapes the signal before the conclusion?
How strong and transferable is the support?
Is this a lesson, reproduction, prototype, benchmark, or documented platform?
What was the dated status screen, in which family and jurisdiction?
Evidence & status legend
Connections can be useful without collapsing unlike records into one claim.
Ask whether the method was independently replicated and whether the conditions match your use.
Reproduction strengthens a specific result; it does not validate every downstream claim.
Existence and implementation are not the same as scientific or clinical validation.
Disclosure is not validation. Status and claim scope require dated, jurisdiction-specific review.
Use it to form a question, not to state product implementation or legal coverage as fact.
State what result would change the current conclusion.
Patent and legal boundary: status labels in this knowledge base are dated research screens, not legal opinions. Expired, lapsed, abandoned, and public-domain candidates can be useful design history only after family, jurisdiction, restoration, continuation, and other rights are checked. Active and pending records are landscape-literacy inputs. Educational or noncommercial use does not automatically clear active patent rights.
Clinical boundary: this site is for research and education. It does not provide diagnosis, treatment, medical-device performance claims, or a substitute for validated clinical instruments and appropriate study design.
A practical thread through the ecosystem
The current PulseSensor is a raw, single-channel, green-light reflectance PPG input with open-source software. It can support careful work on waveform shape, pulse timing, site, contact, sampling, signal quality, motion, and physiological modulation.
It does not measure SpO2 or blood pressure and is not medical equipment. The wider knowledge base places it beside wearable, chip, and clinical ecosystems so accessible experiments can lead to better technical questions.
Worked innovation path
This pilot demonstrates how one topic crosses evidence, patents, products, replication, and future work without claiming that one source proves the next.
Stevens et al. (2024) studied respiratory-rate estimation using off-the-shelf PulseSensors in 15 healthy adults, with useful but limited results.
Read the CC BY paper →Public records describe multiple approaches to respiratory information from pulse-related signals, with unequal status and uncertain claim scope.
Open the patent overlay →PulseSensor supplies a raw green reflectance waveform and open code; it is an accessible experimental input, not a respiratory or medical instrument.
See the implementation record →Predefine windows, rejection rules, metrics, reference timing, and safe breathing conditions. Without a reference, report modulation—not accuracy.
Open the experiment →Publish synchronized PPG and reference signals, code, signal-quality labels, participant-level splits, and failure cases.
See the decision gates →Preserved source collections
The front door indexes these published Pages. Their content, handles, and independent public URLs remain preserved.
The compact three-lane entry: scientific background, PulseSensor use in published research, and a careful experiment.
Open the original hub →Forty-seven papers spanning foundations, respiratory physiology, extraction algorithms, motion, quality, and validation.
Read the bibliography →Seven verified published uses plus signal-chain, open-tool, implementation, limitation, and reproducibility context.
See the research record →Starter protocols, a five-step experiment method, advanced questions, safety boundaries, and reporting guidance.
Choose a project →A respiratory-rate pilot connecting paper, patent, product, replication, open questions, and a benchmark roadmap.
Follow the graph →Wearable, Maxim/Analog Devices chip, and clinical sensor families tied to papers, products, public records, and dated uncertainty.
Explore the atlas →How this index stays trustworthy
The scientific-paper and patent archives are maintained locally with provenance; the public Pages point visitors to primary papers, official records, product documents, and the published source collections.
Prefer primary papers and reviews. Record design, population, site, reference, protocol, results, limitations, and replication state. A citation is not an endorsement.
Record publication and family identifiers, jurisdiction, official source, event/status date, and uncertainty. Refresh before relying on a status label.
Preserve raw data and code when possible. Document hardware, placement, contact, ambient conditions, motion, sampling, preprocessing, reference, exclusions, and failures.