New Non-invasive Screening Tool Shows High Accuracy in CAD Detection

A novel screening test that uses artificial intelligence showed high accuracy in non-invasive coronary artery disease (CAD) detection, according to a new study published in The Journal of Electrocardiology.

Authors state in background to the study that conventional noninvasive diagnostic modalities for at-rest detection of stable CAD are limited by “low sensitivity, and personal expertise.”

The current study compared results of conventional angiographic screening for detection of coronary ischemia (gold standard) with results of a new approach that combines supervised artificial intelligence with 5-lead vectorcardiography, called cardisiography (CSG).

Of the 595 participants in the current multicenter analysis who had been admitted for a coronary angiogram, 62% (n=369) were identified as “affected,” ie, had 1, 2, or 3-vessel disease.  

The primary evaluated outcome was accuracy of the CSG Diagnosis System, validated by a 5-fold nested cross-validation vs angiographic findings.

CSG identified CAD at rest with:

• Sensitivity: 90.2 ±â€¯4.2% for women; 97.2 ±â€¯3.1% for men

• Specificity: 74.4 ±â€¯9.8% for women; 76.1 ±â€¯8.5% for men

• Overall accuracy: 82.5 ±â€¯6.4% for women; 90.7 ±â€¯3.3% for men

"We see great potential in Cardisiography for early detection of symptom-free people with coronary artery disease. Broad-scale early detection has not been possible until now," said co-author Sotirios Spiliopoulos, MD, PhD, Department of Geriatrics, St. Vinzenz-Hospital, in a press release.

Like our content? Show us some love! →Subscribe to ourNewsletter →Like us on Facebook →Follow us on Twitter →Follow us on LinkedIn

Although CAD is the most common type of heart disease-affecting >18 million adults aged ≥20 years-patients often do not have symptoms until the onset of a severe event, making early detection of low blood flow crucial to effective diagnosis, management, and treatment.

Elements of Cardisiography

1. The Cardisiograph: A device that records 3-dimensional electric currents via 5 electrodes applied to the upper body.

2. The Cardisio Cloud: A proprietary artificial intelligence software application that uses multiple algorithms and machine learning to evaluate a patient’s measurements.

The Cardisio Graph collects data from the resting heart for approximately 4 minutes and the results, along with a supplementary electrocardiogram report, are available in just 1 minute.

"Cardisio embodies a new generation of medical diagnostic systems that are mobile and Internet-based, leveraging new Artificial Intelligence capabilities," said Meik Baumeister, co-founder and CEO of Cardisio, GmbH, creator of Cardisio, in the same press release. "The combination of classic cardiology coupled with intelligent high-performance algorithms offers us unprecedented opportunities to change how CAD is managed globally."

Cardisio is currently evaluating partners and studies to support regulatory clearance and deployment in the US and other countries.