For years, camera-based vital sign measurement lived in a regulatory gray area. The technology worked in labs, papers got published, startups raised money on the promise, but nobody had gone through the FDA. That changed in 2025, and the pace of change caught even some insiders off guard.
Two companies secured 510(k) clearances for contactless vital sign software within months of each other. The FDA issued updated guidance that reshapes how wellness and clinical decision support software gets evaluated. And in the EU, MDR enforcement timelines are forcing companies to sort out their classification strategies now, not later. The regulatory picture for camera-based vitals went from ambiguous to rapidly clarifying over the past 18 months.
"FaceHeart Vitals is the first FDA-cleared software development kit solution that enables contactless video-based vital sign measurement via a facial scan on built-in cameras of smartphones within 50 seconds." — FaceHeart Corporation, April 2025
The 510(k) clearances that opened the door
January 2025 was the month camera-based vitals crossed from research curiosity to regulated medical product. PanopticAI, a Hong Kong-based startup spun out of HKUST, received 510(k) clearance for its Vital Signs app. The app uses an iPhone or iPad camera to measure pulse rate without physical contact. It was the first mobile application to receive FDA clearance for contactless pulse rate measurement, and the first Hong Kong company to achieve FDA clearance for a Software as a Medical Device product.
Three months later, FaceHeart Corporation received its second 510(k) clearance. The Taiwanese company had already cleared its FH Vitals SDK for heart rate measurement. The April 2025 clearance (K243966) added respiratory rate, with deviations consistently within plus or minus 2 breaths per minute across tested devices. FaceHeart's SDK can estimate heart rate, respiratory rate, blood pressure, SpO2, and HRV from a 50-second facial scan, though the FDA clearances currently cover only heart rate and respiratory rate specifically.
Both companies took the 510(k) pathway, classifying their products as Class II SaMD. Both relied on remote photoplethysmography (rPPG), extracting physiological signals from subtle facial skin color changes captured by standard cameras. Both process data on-device rather than in the cloud, which addresses latency and privacy considerations simultaneously.
Regulatory pathways for camera-based vital sign software
| Pathway | Jurisdiction | Classification | Requirements | Timeline | Suited for |
|---|---|---|---|---|---|
| FDA 510(k) | United States | Class II SaMD | Predicate device, clinical data, software documentation | 3-12 months | Clinical vital sign claims |
| FDA De Novo | United States | Class I or II (novel) | Risk-benefit analysis, special controls | 6-18 months | Novel devices without predicates |
| General Wellness | United States | Exempt | No disease claims, lifestyle framing | N/A (self-determination) | Fitness and wellness tracking |
| EU MDR CE Mark | European Union | Class IIa or IIb (software) | Notified Body review, clinical evaluation, QMS | 12-24 months | Medical purpose software in EU |
| UKCA | United Kingdom | Aligned with EU MDR tiers | MHRA registration, clinical evaluation | 12-18 months | Post-Brexit UK market |
| PMDA | Japan | SaMD classification per IMDRF | Clinical evaluation, QMS | 12-24 months | Japanese market entry |
The 510(k) pathway worked for PanopticAI and FaceHeart because predicate devices exist for pulse rate and respiratory rate monitoring. The argument is straightforward: the software measures the same physiological parameter as existing cleared devices, just through a different sensing modality. The challenge is clinical validation, since you need to demonstrate equivalence to the predicate's performance.
For vital signs where no predicate exists (say, contactless blood pressure estimation from facial video alone), the De Novo pathway would apply. De Novo is more expensive and slower, but it creates a new classification and product code that subsequent applicants can use as a predicate. Nobody has taken a camera-based De Novo for vital signs yet, but it is likely the route for any company seeking clearance beyond heart rate and respiratory rate.
The 2026 wellness guidance and what it actually says
In January 2026, the FDA released updated guidance for General Wellness products and Clinical Decision Support software. The timing mattered for the camera-based vitals space because many companies position their products for wellness rather than clinical use.
The updated guidance clarified that products intended for lifestyle awareness, habit formation, and self-reflection may fall outside active medical device enforcement when they do not make disease-related claims or drive clinical action. This was not a new concept, but the 2026 version provided more explicit signals around how non-invasive sensing technologies fit into the wellness framework.
The boundary between wellness and medical device remains firm, though. A product likely does not qualify as general wellness if it claims to diagnose, treat, or prevent disease. Referencing clinical thresholds ("your blood pressure is high"), generating medical-style alerts, or recommending treatments all push the software into regulated territory. The same heart rate measurement can fall on either side of this line depending entirely on how the product frames it. Showing someone their resting heart rate trend for fitness awareness is wellness. Flagging tachycardia for clinical follow-up is a medical device.
Commissioner Robert Makary's FDA has leaned toward deregulation in the digital health space during early 2026, but the structural distinction between wellness and medical devices has not changed. Companies still need to make a deliberate choice about where their product sits, and that choice has consequences for claims, labeling, and marketing.
EU MDR and the classification question
The European regulatory picture is different and, for many companies, harder. Under the EU Medical Device Regulation (MDR 2017/745), software with a medical intended purpose is a medical device. Full stop. The MDCG 2019-11 guidance document lays out qualification criteria, and the classification rules in Annex VIII determine which tier a product falls into.
Camera-based vital sign software intended for clinical monitoring will typically land in Class IIa or IIb, depending on the specific vital signs measured and whether the software is intended for diagnosis versus monitoring. Class IIa and above require involvement of a Notified Body for conformity assessment, which adds cost and time.
The standards stack for EU market access is heavy: IEC 62304 for software development lifecycle processes, ISO 13485 for quality management systems, ISO 14971 for risk management, and IEC 62366 for usability engineering. Post-market surveillance under MDR is also more demanding than legacy directive requirements, with mandatory Periodic Safety Update Reports and post-market clinical follow-up.
The practical effect is that CE marking for camera-based vital sign software takes 12 to 24 months and requires a functioning quality management system before you even start the conformity assessment process. For startups used to moving fast, this is often the biggest friction point in European market entry.
What the SaMD framework means for rPPG
The International Medical Device Regulators Forum (IMDRF) published its SaMD framework to harmonize how software is classified across jurisdictions. The framework uses a two-axis matrix: the seriousness of the health condition the software addresses, and the significance of the information the software provides to the healthcare decision.
For camera-based vital sign software, the classification depends on what the software is used for. Measuring heart rate to inform wellness decisions sits at one corner of the matrix. Measuring heart rate to diagnose arrhythmias in a critical care setting sits at the other. Same measurement, same technology, very different regulatory burden.
This is not a theoretical distinction. FaceHeart's clearance covers specific vital signs for specific use contexts. Expanding the intended use to cover additional clinical scenarios or patient populations would require separate regulatory submissions. A company with an rPPG SDK could need different regulatory strategies for the same core technology depending on who uses it and why.
The IMDRF framework also addresses the question of continuous learning. If a camera-based vital sign algorithm updates its model based on new data after deployment, that change may need regulatory review. The FDA's predetermined change control plan concept allows manufacturers to describe anticipated modifications upfront, but the specifics of how this applies to rPPG algorithms that might adapt to new populations or conditions are still being worked out.
What is still missing from the regulatory picture
The cleared products cover heart rate and respiratory rate. That leaves blood pressure, SpO2, HRV, and other parameters that rPPG technology can estimate still without specific FDA clearance through camera-based methods. Blood pressure is particularly significant because it is the vital sign with the largest potential market and the one where regulatory scrutiny is highest, given the clinical consequences of inaccurate readings.
The FDA's approach to blood pressure measurement has historically been conservative. Cuff-based devices go through well-established pathways with clear performance standards. Cuffless blood pressure estimation — whether from PPG wearables or rPPG cameras — faces the challenge of demonstrating accuracy against those established standards using a fundamentally different measurement approach. IEEE 1708a-2019 provides a standard for cuffless blood pressure devices, but applying it to contactless camera-based methods introduces additional variables around distance, lighting, and camera quality that the standard was not specifically designed to address.
There is also no clear regulatory precedent for multi-parameter contactless vital sign panels. A device that simultaneously estimates five vital signs from a facial scan raises questions about how to validate each parameter, how to handle situations where some measurements are reliable and others are not, and how to communicate measurement confidence to users or clinicians.
Where this is heading
The 2025 clearances established that camera-based vital sign software can pass FDA review. That precedent matters more than the specific products involved. Future applicants now have predicate devices for pulse rate and respiratory rate, which lowers the barrier for subsequent 510(k) submissions.
The next 12 to 18 months will likely see additional clearances from companies that have been running clinical studies. The EU will probably see its first CE-marked rPPG medical devices as companies that began the MDR conformity assessment process in 2024 complete their reviews. And the wellness market will continue expanding under the updated General Wellness framework, with companies carefully positioning their products to stay on the right side of the wellness-medical device boundary.
Circadify is developing contactless vital sign estimation technology with regulatory considerations built into the development process from the start. As the regulatory landscape for camera-based vitals continues to take shape, having a clear understanding of where your product sits and what evidence you need is becoming as important as the algorithm itself.
Frequently asked questions
Has the FDA cleared any camera-based vital sign monitoring software?
Yes. PanopticAI received 510(k) clearance in January 2025 for contactless pulse rate measurement via smartphone camera, making it the first FDA-cleared mobile app for this purpose. FaceHeart received two 510(k) clearances — one for heart rate and another in April 2025 for respiratory rate measurement — making its SDK the first FDA-cleared solution covering multiple contactless vital signs.
How is camera-based vital sign software classified by the FDA?
Camera-based vital sign software is generally classified as Software as a Medical Device (SaMD) under Class II, requiring 510(k) premarket notification. The classification depends on the intended use: software marketed for clinical vital sign measurement goes through the medical device pathway, while software positioned for lifestyle awareness may qualify under the General Wellness exemption.
What changed in the FDA's 2026 digital health guidance?
The FDA's 2026 update refined the General Wellness framework and Clinical Decision Support guidance. It provided clearer signals that low-risk, non-invasive sensing technologies can support wellness-oriented experiences without automatic classification as medical devices, as long as they avoid disease-related claims and clinical thresholds.
What standards apply to camera-based vital sign software in the EU?
Under the EU Medical Device Regulation (MDR 2017/745), camera-based vital sign software with a medical intended purpose is classified as a medical device and requires CE marking. Applicable standards include IEC 62304 for software lifecycle, ISO 13485 for quality management, ISO 14971 for risk management, and IEC 62366 for usability engineering.
Related Articles
- What is rPPG Technology — The foundational overview of remote photoplethysmography and how cameras extract physiological signals from skin color changes.
- Privacy and Data Security in Camera-Based Health Monitoring — How camera-based vital sign systems handle data protection, a closely related concern for any regulated health product.
- rPPG Accuracy and Clinical Validation Methods — How rPPG accuracy is measured and validated, directly relevant to the clinical evidence requirements discussed in this post.