Data support use of AI to accurately estimate lesion-specific ischemia


Elucid, a developer of AI software to enable cardiovascular disease detection, has announced the publication of new clinical data in the International Journal of Cardiology.

The study, entitled Coronary plaque assessment of Vasodilative capacity by CT angiography, effectively estimates fractional flow reserve demonstrates that fractional flow reserve (FFR) assessment derived from non-invasive plaque morphology using Elucid’s software is highly accurate as compared to invasive FFR.

The team at Medical University of South Carolina (MUSC), Charleston, USA led by Akos Varga-Szemes and Joseph Schoepf showed that Elucid’s AI software can quantitate both vessel structure and histologically-validated plaque composition from a routine CT angiography (CTA). These measurements were fed into Elucid’s computational neural networks to accurately estimate ischaemia.

“The combination of computed tomography-based fractional flow reserve (FFRct) and plaque morphology provides the clearest overview of myocardial ischaemia in order to determine treatment options for a patient. In our study, we set out to confirm that non-invasive plaque morphology-based FFR may be used to accurately estimate lesion-specific ischaemia,” said Varga-Szemes. “Leveraging Elucid’s software, the study demonstrated a strong level of sensitivity and specificity relative to invasive FFR.”

Several large clinical trials have demonstrated that plaque composition is the culprit of most major heart attacks. However, morphology has not been used to evaluate plaque as the cause of ischaemia because diagnostic options are often invasive and less accurate. The majority of patients with suspected enigmatic coronary artery disease are admitted for an invasive procedure, where catheters and wires are used to measure a specific blockage to determine if it is causing angina, which in turn can guide stent placement required to clear the blockage and restore flow.

“Invasive cardiac catheterisation coupled with pressure sensors is currently the standard of care to evaluate coronary plaque as the cause of ischaemia,” said Andrew Buckler, founder and CTO of Elucid. “Use of the non-invasive Elucid software analysis to assess the vasodilative capacity is as accurate or more so than computational fluid dynamics-based simulations when compared to the standard of care invasive measurement in predicting risk of ischaemia, and this is scientifically and clinically significant as it provides a direct measure on the functional significance of the lesion where it actually originates. Elucid’s unique combination of objective validation and AI processing is a game-changer for measuring both FFR and plaque stability, helping physicians make treatment decisions requiring only a non-invasive test.”


Please enter your comment!
Please enter your name here