Tempus AI publishes multi-site validation of ECG-AF tool in Heart Rhythm journal

Tempus AI's FDA-cleared ECG algorithm accurately predicted one-year atrial fibrillation risk in a study of over 4,000 patients aged 65 and older, published in Heart Rhythm. The peer-reviewed results bolster the company's push to convert its AI tools into reimbursable diagnostic revenue.

Categorized in: AI News Healthcare
Published on: Jun 22, 2026
Tempus AI publishes multi-site validation of ECG-AF tool in Heart Rhythm journal

Tempus AI's FDA-cleared ECG-AF algorithm accurately predicted one-year atrial fibrillation or flutter risk across more than 4,000 patients aged 65 and older, according to peer-reviewed results published in Heart Rhythm. The multi-site validation study gives the company fresh clinical ammunition as it works to convert its growing library of algorithms into reimbursable diagnostic revenue.

The study confirms the software's real-world performance outside the controlled conditions of its original regulatory submission. For a company still navigating unprofitable quarters and rich sales multiples, peer-reviewed evidence like this matters - it speaks directly to the clinical credibility payors and health systems demand before adopting AI-driven decision support tools.

The clinical data behind the announcement

The research drew from a cohort of over 4,000 patients, all aged 65 or older, and tested the algorithm's ability to forecast atrial fibrillation or flutter events within a one-year window. Publication in a specialty cardiology journal signals that the findings met the scrutiny of independent reviewers, which carries more weight with clinicians than a company press release alone.

Tempus AI already held FDA clearance for the ECG-AF tool. This study adds a layer of post-market evidence that the algorithm performs as intended across multiple clinical sites, not just the datasets used during the clearance process.

Where this fits in Tempus AI's commercial push

The ECG-AF validation arrives alongside the rollout of Tempus's next-generation Lens platform, which packages the company's multimodal AI capabilities for 19 of the top 20 biopharma clients. Together, they illustrate a deliberate strategy: turn validated algorithms into repeatable software and data contracts.

Investors watching the stock are tracking whether these tools can offset the company's high R&D and capital expenditure needs. The bullish analyst case already projects revenue reaching roughly $2.5 billion with positive earnings by 2029. Clinical validations like this one support that narrative, but the critical variable remains reimbursement. If payors move slowly on covering AI-based clinical decision support, even a growing portfolio of FDA-cleared algorithms won't guarantee commercial traction.

Contrasting views on the investment case

The spread between optimistic and cautious projections for Tempus AI is wide. Some analysts see a path to profitability within five years, while others point to the uncertainty around how quickly health insurers will reimburse AI-driven diagnostics. The ECG-AF study provides a concrete data point for both sides: it strengthens the clinical argument without resolving the reimbursement question.

"The new multi-site validation of its FDA-cleared ECG-AF tool adds clinical weight to its cardiovascular AI story, but does not, by itself, change that the key near term catalyst remains broader monetization of AI algorithms," the analysis noted.

Why this matters for healthcare professionals

For clinicians and health system leaders, the study offers independent evidence that an AI tool can stratify arrhythmia risk in older patients - a population where early detection changes treatment decisions. Peer-reviewed validation in a journal like Heart Rhythm is the type of signal many healthcare organizations look for when building the case to pilot or deploy AI-enabled diagnostics. The reimbursement landscape remains uncertain, but the clinical evidence base is getting harder to ignore.


Get Daily AI News

Your membership also unlocks:

700+ AI Courses
700+ Certifications
Personalized AI Learning Plan
6500+ AI Tools (no Ads)
Daily AI News by job industry (no Ads)