DEEPVESSEL FFR
KeyaMed NA Inc.DEEPVESSEL FFR is a clinical software tool that uses deep learning to analyze previously acquired coronary CT angiography (CTA) images. It generates three-dimensional models of coronary artery trees and estimates fractional flow reserve (FFR) values to help clinicians assess the functional severity of coronary artery disease. It supports doctors by providing additional insight beyond anatomical imaging, using AI-based physiological simulation to improve diagnosis and treatment planning for heart vessel conditions.
Intended Use
DEEPVESSEL FFR is a coronary physiological simulation software for the clinical quantitative and qualitative analysis of previously acquired Computed Tomography DICOM data for clinically stable symptomatic patients with coronary artery disease. It provides DVFFR (a CT-derived FFR measurement) computed from static coronary CTA images using deep learning neural networks that encode imaging, structural, and functional characteristics of coronary arteries through learning. DEEPVESSEL FFR analysis is intended to support the functional evaluation of coronary artery disease. The analysis results are used by qualified clinicians in conjunction with clinical history and other diagnostic tests.
Technology
DEEPVESSEL FFR uses deep learning neural networks to process static coronary CTA images, generating a 3D coronary artery tree model via segmentation, which may be manually corrected. A multi-layer perceptron network and a bidirectional recursive neural network estimate semi-continuous FFR values along the coronary artery centerlines. Outputs include a detailed PDF report and 3D model storing the FFR values mapped to the artery surface.
Performance
Performance testing included software verification and validation, human factors studies for safe use, reproducibility and repeatability evaluations on diverse CT scans, and a multi-center clinical validation study involving 244 patients. The clinical study showed per-vessel sensitivity of 86.9% and specificity of 86.7%, exceeding the predefined targets. Patient-level diagnostic accuracy was 85.2%, with positive predictive value of 79.6% and negative predictive value of 90.1%, demonstrating good diagnostic performance and effectiveness.
Predicate Devices
No predicate devices specified
Device Timeline
- 1
Submission
11/19/2021
4 months - 2
FDA Approval
4/1/2022
Ready to Sharpen Your Edge?
Join hundreds of your peers who rely on RadAI Slice. Get the essential weekly briefing that empowers you to navigate the future of radiology.
We respect your privacy. Unsubscribe at any time.