A light shined on a patient's toe, a sensor measuring the reflection, an AI algorithm processing the result—all in seconds. Researchers at UC San Diego have discovered that this simple photoplethysmography (PPG) scan can detect peripheral artery disease with 83% accuracy, potentially transforming how millions of Americans at risk are screened for a life-altering condition.
Peripheral artery disease, a buildup of plaque in blood vessels that restricts blood flow to the legs, affects 8 to 12 million Americans and disproportionately impacts marginalized communities. It's a leading cause of limb amputation, yet it often goes undetected until patients experience serious complications. The reason is partly practical: diagnosing PAD currently requires a specialized clinic visit for an ankle-brachial index (ABI) test that takes 15 to 30 minutes to complete—a significant barrier for patients without reliable transportation or financial resources.
The breakthrough came from an unlikely conversation. Dr. Mattheus Ramsis, an assistant professor of medicine and medical director of cardiology informatics at UC San Diego School of Medicine, learned from vascular surgeon Dr. Elsie G. Ross that PPG recordings are routinely taken on the toe during ABI testing anyway. "The lightbulb went off for me at that moment," Ramsis recalls. If that data was already being collected, why not harness it?
Working with Ph.D. student Ava J. Fascetti and fourth-year medical student Mustafa H. Naguib in the lab of Edward J. Wang, Ramsis assembled a dataset of more than 10,000 toe PPG recordings from over 3,500 patients who underwent ABI tests at UC San Diego Health between 2020 and 2025. The team extracted 78 PPG features correlated with patients' ABI results and developed a machine learning model to detect PAD based solely on that light-reflection data. The results were striking: the model correctly identified PAD cases approximately 83% of the time, substantially outperforming the 60–65% accuracy achieved using traditional clinical risk-factor assessment alone. Adding a patient's smoking status boosted accuracy by another 2%.
What makes this finding particularly significant is its equity dimension. The model performed similarly across Black, Hispanic, and white patients, and held up across two separate UC San Diego Health campuses, even in patients with coronary artery disease, diabetes, and end-stage renal disease. This consistency suggests the tool could work reliably across diverse populations, addressing a critical gap in screening equity.
A PPG screen takes only a few seconds compared to 15–30 minutes for an ABI test. Because 95% of people own a smartphone or some advanced device, Ramsis envisions screening becoming accessible at the point of care in clinics—or even at home through a smartphone app. High-risk patients could monitor themselves without waiting for a clinic appointment, potentially catching PAD before it progresses to limb-threatening stages.
The team's next step is to validate the approach across multiple PPG-capturing devices, including smartphones, pulse oximeters, and wearables, to ensure the model works in real-world settings. While ABI testing won't be replaced, PPG screening could complement it, removing the transportation, financial, and institutional barriers that currently limit access to early detection. For millions of Americans living with undiagnosed PAD, that seconds-long scan could mean the difference between losing a leg and keeping it.