FDA Clears Tempus's AI Tool for Pulmonary Hypertension

FDA granted 510(k) clearance on 24 August 2026 to Tempus ECG-PH, an AI-enabled software device that reads a standard 12-lead ECG for signs of pulmonary hypertension. It is the third FDA clearance in Tempus AI's cardiovascular algorithm portfolio, following earlier clearances for atrial fibrillation risk and low ejection fraction detection.

What the clearance actually covers

Tempus ECG-PH takes a resting standard 12-lead ECG recording as its input and outputs a binary signal: does this tracing show patterns associated with elevated mean pulmonary artery pressure above 20 mmHg, the hemodynamic threshold used to define pulmonary hypertension. FDA's clearance is explicit about the device's limits. It is not a stand-alone diagnostic — a positive flag is a prompt for further workup, not a diagnosis in itself — and it is not validated for serial monitoring of a patient already being treated, or for ECGs with paced rhythms, which alter the waveform in ways the algorithm was not trained to interpret. Clinicians reading a flagged result are expected to weigh it against the underlying ECG, the patient's symptoms, other diagnostic information, and clinical history, not act on the software output in isolation.

A third clearance, not a first

Tempus ECG-PH extends a pattern rather than starting one: it is the company's third FDA-cleared cardiovascular algorithm, following clearances for Tempus ECG-AF, which flags atrial fibrillation risk, and Tempus ECG-Low EF, which flags reduced ejection fraction. All three share the same basic architecture — an AI model trained to extract a specific risk signal from a widely available, low-cost input (a standard 12-lead ECG) that most patients already have on file or can obtain quickly. For a device maker, a repeat 510(k) pathway on the same underlying data type is a materially different regulatory lift than a first-of-kind submission: FDA has an established predicate and review framework for this device category, even as each new indication requires its own clinical validation.

The clinical case for ECG-based screening

Brandon Fornwalt, MD, PhD, Tempus's Senior Vice President of Cardiology, framed the clearance around a diagnostic gap: pulmonary hypertension is underdiagnosed, he said, because its early symptoms are non-specific and the gold-standard confirmatory test — right heart catheterization — is invasive and not something clinicians order without a strong prior suspicion. An ECG-based screening flag, if it performs as validated in real-world use, offers a cheaper and non-invasive way to surface candidates for that confirmatory workup earlier in the disease course, before symptoms progress to the point that clinical suspicion is already high. Whether that promise holds outside the clearance's validation population is a question for post-market performance, not the clearance itself, which addresses only the device's authorized use, not real-world diagnostic yield.