The U.S. Food and Drug Administration (FDA) has granted 501(k) clearance for a first-of-its-kind pulmonary hypertension (PH) algorithm. Anumana, a health technology company, developed the artificial intelligence (AI)-enabled software-as-a-medical-device (SaMD) to detect early signs of PH using standard 12-lead electrocardiograms (ECGs).
PH is a progressive, life-threatening disease that is often difficult to diagnose due to nonspecific early symptoms, like dyspnea. Diagnostic delays, which frequently exceed two years, lead to increased morbidity and mortality, according to an Anumana press release on the FDA approval.
The new algorithm augments ECGs by helping clinicians identify subtle, irregular patterns associated with PH that may not be visible to the human eye. According to Anumana, the tool integrates with EHR systems and operates within the technology network without patient data transference. This enhancement can flag patients who require follow-up testing and allow for improved surveillance or early diagnosis and treatment.
“Pulmonary hypertension is often difficult to recognize until it has progressed significantly, leaving patients and physicians at a disadvantage,” said Paul Friedman, MD, chair of the department of cardiovascular medicine at Mayo Clinic and a member of the Anumana advisory board. “FDA clearance of Anumana’s PH algorithm provides clinicians with a practical solution to identify PH earlier and determine appropriate next steps in care, expanding the clinical utility of routine ECGs.”
Researchers used more than 250,000 de-identified patient records from Mayo Clinic to train the ECG-AI PH algorithm. This work was detailed in the paper, “An Electrocardiogram-Based AI Algorithm for Early Detection of Pulmonary Hypertension,” published in European Respiratory Journal. In an independent, multicenter study of 21,066 patients across five U.S. health systems, the algorithm helped detect PH with 73% sensitivity and 74.4% specificity in adult patients who presented with dyspnea.
In a real-world analysis study of patients with an ECG available between symptom onset and PH diagnosis, it identified more than 85% of patients with pulmonary arterial hypertension (PAH) and 78% of patients with chronic thromboembolic pulmonary hypertension (CTEPH). These results support the algorithm’s use in early detection of these two treatable PH subgroups, Anumana said.
“As the first PH algorithm cleared for use with standard 12-lead ECGs, it is broadly accessible across care settings, integrates directly into existing clinical workflows, and supports clinical decision-making in real time, with the potential to help identify patients earlier in their disease course,” said President and COO of Anumana Simos Kedikoglou.
