Point of care device predicts PH in patients with new-onset symptoms

CorVista System may enable earlier diagnosis of progressive disease: Study

Margarida Maia, PhD avatar

by Margarida Maia, PhD |

Share this article:

Share article via email
A speaker points to a white board with data while leading a meeting at a conference.

The CorVista System, a point of care device that applies a machine learning algorithm to decode data, can predict the presence of pulmonary hypertension in patients with new-onset symptoms of cardiovascular disease, a study shows.

The device’s algorithm performed about as well as transthoracic echocardiography, or TTE, a standard test that uses high-frequency sound waves to measure blood flow through the heart, according to CorVista Health, which developed the CorVista System and conducted the study.

Data from that study was presented in a poster, titled “Machine Learning to Detect Pulmonary Hypertension at Point-of-Care,” at this year’s American Heart Association (AHA) annual meeting, held Nov. 11-13 in Philadelphia.

The U.S. Food and Drug Administration (FDA) last year granted breakthrough device designation to a CorVista System add-on intended to calculate the likelihood of high blood pressure in the lungs’ blood vessels.

“[These] important data [demonstrate] the feasibility to detect pulmonary hypertension, as well as enabling potentially earlier diagnosis in the progression of the disease,” Charles Bridges, MD, executive vice-president and chief scientific officer of CorVista, said in a company press release. Bridges gave the presentation at the AHA meeting.

Recommended Reading
An illustration of people looking at information shown on a tablet's screen.

EchoRight Pro app aims to speed pulmonary hypertension diagnosis

CorVista says point of care device can aid underserved patients

Diagnosing pulmonary hypertension can be difficult because many of its symptoms overlap with those of other conditions. The process often is lengthy and involves several different tests.

Moreover, existing diagnostic tools, even TTE, can be insufficient.

For example, tricuspid regurgitation velocity (TRV), the most significant parameter to assess pulmonary hypertension on an echocardiogram, cannot be measured in nearly half of patients, making TTE less accurate.

With the CorVista point of care device, signals naturally emitted by the heart are collected while the patient is at rest. A machine learning algorithm runs its analysis of the data on the cloud (a remote data storage and processing service). The results of the analysis become available to doctors in a secure web portal shortly after the procedure is completed.

The procedure is rapid and noninvasive, and, according to the company, works “without the use of radiation, contrast agents, injections, fasting or exercise.”

We believe the CorVista System can make a tremendous impact to patients suffering from pulmonary hypertension, especially in rural and underserved populations.

In the new study, done in collaboration with researchers at Canada-based Analytics 4 Life and the University of Michigan, signals were captured using the investigational CorVista System in newly symptomatic patients at rest for 3.5 minutes.

“We believe the CorVista System can make a tremendous impact to patients suffering from pulmonary hypertension, especially in rural and underserved populations,” Bridges said.

The researchers also collected the results of TTE within 90 days, or right heart catheterization within seven days. Right heart catheterization, an invasive procedure, measures pressure inside the pulmonary arteries, which are the blood vessels that supply the lungs.

A machine learning algorithm was trained to identify features that can distinguish high mean pulmonary arterial pressure, known as mPAP.

The algorithm was able to discriminate patients with no evidence of diastolic dysfunction — when the heart’s low left chamber does not relax as it should — nor pulmonary hypertension on TTE from those with a mPAP equal to or greater than 25 millimeters of mercury (mmHg) on right heart catheterization. 

It did so with a sensitivity of 87% and a specificity of 83%. Sensitivity refers to how well the algorithm can predict pulmonary hypertension in patients who actually have it, whereas specificity is the proportion of patients who test negative among those who do not have it.

In a subgroup of patients with pre-capillary pulmonary hypertension, where high blood pressure affects the pulmonary arteries before reaching the small blood vessels in the lungs, sensitivity was 92% and specificity 83%.

“We are pleased to present [these] important data at this year’s AHA meeting,” said Don Crawford, president and CEO of CorVista.

“Pulmonary hypertension is a challenging condition to diagnose, especially early in the disease progression,” Crawford noted.

The researchers noted that PH typically is diagnosed “years after symptom onset.”

“An algorithm with performance comparable to TTE can be developed to assess the likelihood of [pulmonary hypertension] in patients with new onset symptoms of cardiovascular disease,” the team wrote.

“Performance is preserved in all subgroups, including pre-capillary [pulmonary hypertension], and is not limited by the availability of TRV,” they concluded.

The FDA-cleared CorVista System with a coronary artery disease add-on has the ability to non-invasively indicate the likelihood of coronary artery disease at the point of care, within minutes, as an aid in diagnosis, according to the team.

A Conversation With Rare Disease Advocates