Way of Diagnosing PH in Common Heart Test Gets FDA Support

Lindsey Shapiro, PhD avatar

by Lindsey Shapiro, PhD |

Share this article:

Share article via email
algorithm for diagnosis of PH | Pulmonary Hypertension News | news announced illustration

Anumana’s electrocardiogram (ECG)-based algorithm for the early detection of pulmonary hypertension (PH) has been designated a breakthrough device by the U.S. Food and Drug Administration (FDA).

Developed through a collaboration of scientists at Anumana, Janssen Research and Development, and the Mayo Clinic, the artificial intelligence (AI)-powered algorithm is designed to enable standard heart tests using ECG to quickly and precisely predict a person’s likelihood of having PH.

The technology could help with the significant unmet need for earlier PH diagnoses and treatment start, leading to better outcomes, the company reported in a press release.

Recommended Reading
pulmonary arterial hypertension mortality | Pulmonary Hypertension News | illustration of thumb's down

PAH Mortality Risk ‘Unacceptably High’ in the Modern Era, Study Finds

Breakthrough device designation is given to medical equipment with the potential to more effectively diagnose or treat life-threatening diseases. It’s intended to speed the algorithm’s development and its regulatory review for possible approval.

“The FDA’s Breakthrough Device Designation for Anumana’s PH Early Detection Algorithm is one step forward for the field of ECG AI overall, and more saliently, a giant leap forward for PH patients,” Venky Soundararajan, PhD, co-founder and chief scientific officer of Anumana and its parent company, nference, said in the release.

PH can be accompanied by non-specific symptoms, like shortness of breath, causing it to go unnoticed until the disease has progressed. Delays in diagnosing PH are common, the company noted, often taking more than one year.

“While therapeutic options for patients with pulmonary hypertension have evolved in recent years, we have not seen significant advancement in reducing the time from symptom onset to diagnosis — and our hypothesis was that data science could help change this,” said Najat Khan, PhD, chief data science officer and global head of strategy and operations at Janssen.

The research team leveraged the fact that many primary care and emergency room settings are equipped for ECG — a noninvasive test of the heart’s electrical activity. It’s among the first diagnostic tests used to analyze changes in heart structure and workings that characterize PH.

“Electrophysiology waveforms hold immense untapped potential for detecting diseases earlier in their natural history, particularly for conditions in which earlier diagnosis and therapeutic intervention can prolong survival and improve quality of life,” Soundararajan said.

Anumana’s AI-powered algorithm is designed to enhance the predictive power of these ECG recordings. During a person’s ECG recording, the algorithm will use data from the nference platform, which contains information from more than 6 million de-identified patient records and over 8 million ECG recordings, to analyze the ECG and, reportedly “within seconds,” predict the likelihood of PH.

While still in development and testing, Anumana believes the algorithm has the potential to transform and accelerate PH diagnoses, leading to better outcomes. The algorithm will be available to physicians as downloadable software to a smartphone, tablet or computer, or available through a Cloud-based source.

“Early diagnosis of pulmonary hypertension is paramount due to its progression and potential severity,” said Paul Friedman, MD, chair of the Department of Cardiovascular Medicine at the Mayo Clinic.

“The addition of AI to a standard ECG — a painless, inexpensive, widely used test that is routinely performed — transforms the ECG into a screening tool for PH, with the opportunity to improve outcomes via early detection by guiding appropriate testing,” Friedman added.

Anumana was launched by nference in 2021 with support from the Mayo Clinic Platform to develop AI-enabled algorithms.