CardioMe's Non-Invasive Heart Attack Predictor Clears FDA: A Blood Flow Sensor You Wear for 30 Seconds a Day

The Problem With Cardiac Monitoring Today

Cardiovascular disease remains the leading cause of death globally, killing approximately 20 million people annually. The tragedy is that most of these events are not random. Heart attacks are, in the majority of cases, the culmination of a gradually developing coronary artery blockage that a properly timed intervention could resolve. The challenge has always been detection: existing diagnostic tools—CT angiography, nuclear stress tests, cardiac catheterization—are expensive, require clinical settings, and carry their own risks. Wearable heart monitors exist, but they measure electrical signals (ECG), not the hemodynamic changes that precede a blockage-induced event.

CardioMe, a four-year-old medical device company headquartered in Tel Aviv, believes it has solved this detection gap. On November 9th, the company announced that its flagship product, the PalmSense C-100, has received FDA 510(k) clearance for use as a non-invasive predictor of adverse cardiac events in adults with suspected coronary artery disease.

The Technology: Ultrawideband Pulse Radar + AI

The PalmSense C-100 is a handheld device roughly the size of a TV remote. To use it, a patient holds the device against their bare chest for 30 seconds while the device emits a low-power ultrawideband (UWB) radar pulse that penetrates approximately 8 centimeters into thoracic tissue. The reflected signal is processed by CardioMe's proprietary PulseAI algorithm running on a dedicated neural accelerator chip.

What PulseAI is looking for is not electrical activity but hemodynamic signatures—patterns in the pulse echo that indicate turbulent blood flow in the coronary arteries, a tell-tale sign of stenosis (narrowing). The algorithm was trained on a dataset of 340,000 de-identified UWB pulse recordings paired with corresponding coronary CT angiography results, making it, CardioMe claims, the largest such training dataset in existence.

In the FDA-submitted pivotal trial, PalmSense achieved:

• 91.4% sensitivity for detecting hemodynamically significant coronary artery blockage (≥70% luminal narrowing)
• 87.2% specificity for ruling out significant blockage
• A negative predictive value of 96.3%—meaning a "clear" reading was correct 96.3% of the time

The Clinical Workflow

The device is not intended to replace clinical diagnostics. Instead, CardioMe is positioning PalmSense as a triage and monitoring tool that enables cardiologists to identify high-risk patients between scheduled appointments. The recommended protocol: a 30-second scan each morning before breakfast. If the device detects a concerning hemodynamic pattern, it prompts the user (via Bluetooth-connected smartphone) to contact their cardiologist within 24 hours.

Cardiologists receive a PDF report summarizing the patient's hemodynamic trend over the preceding 7 days, allowing them to see whether a pattern is acute or chronic.

Cost and Accessibility

The PalmSense C-100 is priced at $890 per unit for consumers, with a companion subscription service at $29/month that includes AI model updates, cardiologist report delivery, and a battery replacement program. The company is pursuing insurance reimbursement coding through CMS and has announced partnerships with three major US pharmacy chains for in-store demos and distribution.

What This Means for Cardiac Care

If CardioMe's claims hold up in broader post-market use, the clinical implications are significant. The ability to detect coronary blockage patterns in a 30-second, non-invasive scan at home—without needles, without a clinic visit, without radiation—could shift cardiac risk management from reactive to proactive. Cardiologists who reviewed the FDA submission data (presented at the ACC Advanced Cardiovascular Imaging conference in October) called the results "compelling" while noting the need for broader real-world validation.

The first PalmSense units are expected to ship to US consumers in February 2028.