The short answer is no. Your phone cannot detect a heart attack before it happens, and no app on the market today can predict an acute myocardial infarction. If you're having chest pain, shortness of breath, or other symptoms, put the phone down and call 911. That hasn't changed, and it won't change for a while.
But the longer answer is more interesting. Smartphones are getting surprisingly good at measuring things that relate to heart health: resting heart rate, heart rate variability, and even irregular rhythms like atrial fibrillation. None of these are the same as predicting a heart attack, but they're the kind of data points that, tracked over time, might eventually contribute to earlier warning systems. The gap between what phones can do today and what people hope they'll do tomorrow is worth understanding clearly.
"Average heart rate throughout the day has been shown to be a strong independent predictor, even more so than resting heart rate, for all-cause mortality." — Korshøj et al., cited in Poh et al., Google Research (2025)
What smartphones can actually measure right now
Your phone has a camera and a flashlight. That's enough hardware to measure your heart rate. The technique is called photoplethysmography (PPG): when your heart beats, blood volume changes slightly in your skin, altering how light is absorbed and reflected. A camera can pick up those tiny color shifts, frame by frame, and extract a pulse signal.
There are two versions of this. Fingertip PPG requires you to press your finger against the camera lens and flash. It's been around for years and works reasonably well. The more recent approach, remote photoplethysmography (rPPG), skips the touching part entirely. The front-facing camera just looks at your face during normal use and reads the signal from there.
A 2025 study from Google Research put rPPG through its most rigorous real-world test yet. Di Achille, Wu, Borac et al. built a system called PHRM and validated it on 185,970 videos from 205 participants in both lab and free-living conditions. The results: heart rate measurement error below 10% across light, medium, and dark skin tones, with no statistically significant accuracy gap between skin tone groups. Daily resting heart rate estimates came within 5 bpm of a wearable tracker (Di Achille et al., 2025).
That's a meaningful result. It means a phone you're already looking at can passively measure your heart rate without you doing anything special. No wrist strap, no finger on the lens, no dedicated measurement session.
Beyond basic heart rate, smartphones can also derive heart rate variability (HRV) from PPG signals. HRV matters because reduced variability is associated with increased cardiovascular risk. A review in PMC noted that low HRV is an independent predictor of increased mortality after acute myocardial infarction and in patients with chronic heart failure (PMC, 2023). Tracking HRV over weeks or months could theoretically flag a downward trend worth discussing with a doctor.
What smartphones cannot do
Here's where the gap gets wide. A heart attack happens when a coronary artery becomes blocked, cutting off blood supply to heart muscle. Detecting that requires imaging (like a coronary angiogram), blood tests (troponin levels), or at minimum a 12-lead ECG showing ST-segment changes. A phone camera reading skin color changes from your face is not going to catch a plaque rupture in a coronary artery. The physiology just isn't there.
Smartphones also cannot reliably detect most arrhythmias beyond atrial fibrillation. Ventricular tachycardia, heart blocks, and ST-elevation patterns that signal an active heart attack require electrical measurement that PPG doesn't provide. PPG tells you about pulse timing and regularity. It tells you almost nothing about the electrical conduction pathways inside the heart.
The FDA has cleared certain smartwatch ECG features for atrial fibrillation detection specifically. Samsung received FDA clearance in July 2024 for its Irregular Heart Rhythm Notification feature on Galaxy Watch devices. But these clearances are narrow: they cover AFib screening in adults over 22, not heart attack prediction, not general arrhythmia detection. The regulatory bodies are being careful about what these devices can claim, and for good reason.
Comparing smartphone cardiac monitoring capabilities
| Capability | Fingertip PPG (Phone Camera) | Facial rPPG (Front Camera) | Smartwatch ECG | Clinical 12-Lead ECG |
|---|---|---|---|---|
| Heart rate measurement | Yes, well validated | Yes, validated in large studies | Yes | Yes |
| Heart rate variability | Yes, with good signal | Emerging, less validated | Yes | Gold standard |
| Atrial fibrillation screening | Yes, 94% pooled sensitivity | Under investigation | FDA-cleared for some devices | Gold standard |
| Heart attack detection | No | No | No (single-lead limitations) | Yes (ST-segment analysis) |
| Continuous passive monitoring | No (requires finger placement) | Yes (during normal phone use) | Yes (worn on wrist) | No (clinical setting only) |
| Skin tone equity validated | Limited data | Yes (Poh et al., 2025) | Limited data | N/A |
| Regulatory clearance | Few apps cleared | None yet | Select devices cleared | Established |
Sources: Di Achille et al. (2025), Jamaladin et al. (2022), Samsung FDA submission K240909.
The table makes the tradeoffs visible. Fingertip PPG on phones is the most validated mobile approach but requires active participation. Facial rPPG is the most passive but least regulated. Smartwatch ECG sits in between. And none of them replace what a clinical ECG can tell a cardiologist about an acute cardiac event.
Atrial fibrillation: where phone screening actually works
If there's one cardiac condition where smartphones show genuine promise, it's atrial fibrillation. AFib affects between 2.7 million and 6.1 million Americans and is a major risk factor for stroke. It's also intermittent in many patients, making it easy to miss during occasional doctor visits.
A 2022 systematic review and meta-analysis by Jamaladin et al. in BMJ Open examined 28 studies comparing smartphone PPG against reference ECG for AFib detection. Across 11,404 participants (2,950 in AFib), pooled sensitivity was 94% (95% CI: 92-95%) and specificity was 97% (96-98%). Those numbers look impressive, but the authors were blunt about limitations: most studies were small, conducted in controlled clinical settings, and potentially subject to publication bias. None met all quality criteria on the QUADAS-2 assessment tool.
Real-world performance will almost certainly be lower. People don't hold their phones as still as study participants do. Ambient lighting varies. Some people have cold fingers or nail polish that degrades the signal. The gap between clinical trial accuracy and living room accuracy is a recurring theme in mobile health research.
Still, even imperfect screening has value when the alternative is no screening at all. Most AFib goes undiagnosed until a stroke or other complication forces the issue. If a phone app catches an irregular rhythm and prompts someone to see a cardiologist, that's a net positive, as long as people understand what the app is and isn't telling them.
The rPPG angle: monitoring without trying
What makes rPPG particularly interesting for long-term cardiac monitoring is the passivity. You don't have to remember to take a measurement. You don't have to strap anything on. You just use your phone normally, and the front camera picks up your pulse from your face while you're reading the news or scrolling through messages.
The Google Research PHRM study is the largest prospective validation of this approach to date, with over 400,000 total videos across development and validation sets. The system worked during everyday smartphone use, not in a controlled lab with perfect lighting and a chin rest. That's the difference between a proof of concept and something that could actually scale.
For rPPG to become clinically meaningful, two things need to happen. First, accuracy needs to improve for derived metrics beyond basic heart rate. Respiratory rate estimation from facial video is still unreliable in most implementations. Blood oxygen saturation via camera remains largely unproven outside controlled settings. Second, regulatory frameworks need to catch up. No rPPG-based system has FDA clearance for any cardiac measurement. That's a significant barrier to clinical adoption, even if the research results are promising.
Circadify is developing contactless vital sign measurement technology using rPPG. The company's camera-based approach is built to extract heart rate and other physiological signals from standard video, without requiring physical contact or dedicated hardware. This kind of passive monitoring could eventually contribute to longitudinal cardiovascular health tracking, though the technology is still maturing across the industry.
Where this is heading
The trajectory is fairly predictable. Heart rate measurement via smartphone camera is already accurate enough for consumer health tracking. AFib screening is approaching clinical utility but needs better real-world validation and clearer regulatory pathways. Heart attack prediction remains far out of reach for any consumer device.
The more realistic near-term scenario is this: your phone tracks your resting heart rate and HRV passively over months and years. If those trends shift meaningfully, the system flags it and suggests a doctor visit. That's not predicting a heart attack. It's catching a change in baseline cardiovascular function that might warrant further investigation. It's less dramatic than the headline implies, but potentially more useful.
The research pipeline includes larger and more diverse validation studies, better algorithms for respiratory rate and SpO2 from camera, and work on making rPPG robust to real-world conditions like variable lighting, movement, and cosmetics. Getting the basics reliable across all skin tones and usage contexts matters more than chasing headline-friendly claims about predicting cardiac events.
If you searched "can my phone detect a heart attack," the honest answer is: not today, and probably not soon. But your phone can tell you things about your heart that you couldn't easily know five years ago. That's worth something, as long as you don't mistake screening for diagnosis, or a heart rate reading for a clean bill of health.
Frequently asked questions
Can my smartphone actually detect a heart attack?
No. Smartphones cannot detect a heart attack (acute myocardial infarction) as it happens. What they can measure is heart rate, heart rate variability, and in some cases irregular rhythms like atrial fibrillation. These are risk indicators, not diagnostic tools for heart attacks. If you suspect a heart attack, call emergency services immediately.
How does a phone camera measure heart rate?
Smartphone cameras measure heart rate using photoplethysmography (PPG). When you place your finger on the camera lens, or when the front-facing camera captures your face, subtle changes in skin color caused by blood flow are detected and analyzed to determine pulse rate. This technique is called remote photoplethysmography (rPPG) when no physical contact is required.
How accurate is smartphone heart rate monitoring?
Accuracy varies by method and conditions. A 2025 study from Google Research involving over 400,000 videos found that passive rPPG heart rate monitoring during normal smartphone use achieved less than 10% mean absolute percentage error across light, medium, and dark skin tones. Fingertip PPG apps tend to be more accurate than camera-only facial rPPG, but both are improving.
Can phones detect atrial fibrillation?
Research suggests smartphone PPG can detect atrial fibrillation with pooled sensitivity of 94% and specificity of 97%, according to a 2022 systematic review of 28 studies. However, the review authors noted that most studies were small, conducted in clinical settings, and may overestimate real-world performance. Smartphone-based AF screening is not a replacement for clinical ECG diagnosis.