
DEEPVESSEL FFR is FDA Cleared, CE Marked, and NMPA Approved.

Improving Diagnosis & Treatment of Patients with Coronary Artery Disease with CT FFR
DEEPVESSEL FFR ( DVFFR) is a software medical device that uses deep learning technology to perform a non-invasive physiological functional assessment of the coronary arteries using coronary computed tomography angiography. The software processes coronary CTA images semi-automatically, generates a three-dimensional model of the coronary artery tree, and computes non-invasive DVFFR values. A PDF report of the derived information is sent electronically to physicians. DEEPVESSEL FFR is intended to support the functional evaluation of coronary artery disease.
DEEPVESSEL FFR applies DeepVessel and DeepFFR techniques, Keya Medical’s proprietary deep learning technologies built on the latest advances in computer vision and medical image analysis. This new approach gives a strong performance advantage with fast processing speed and accurate non-invasive FFR analysis.
How it Works

Clinician sends coronary CTA DICOM images to Keya Medical

Deep learning algorithms and Keya Medical trained analysts perform the DeepVessel FFR analysis

An interactive report and 3D model of the coronary arteries is sent electronically to the clinician for review

Product Features
Non-invasive functional assessment from CTA scan
Enables accurate and fast FFR analysis
Provides both anatomical and functional information
Novel coronary CT FFR technique driven by AI
Reduces medical cost
FDA cleared
CE marked
Advantages

Non-Invasive
Calculates non-invasive DVFFR values from coronary CTA images to reduce unnecessary coronary catheterizations

Accurate
Achieves 90% diagnostic accuracy using invasive FFR as the ground truth

Comprehensive
Provides both anatomical and functional information to improve clinical decision making

Fast
Delivers results to clinicians quickly for timely decision making

Accessible
Results can be viewed on mobile, PC, and tablet devices

Economical
Reduces costs from invasive procedures and unnecessary therapies
DeepVessel FFR Diagnostic Performance

Clinical Validation of DEEPVESSEL® FFR: AI CT-FFR Analysis by Keya Medical
The ADAPT Study: Assessment of the DiAgnostic Performance of DeepVessel FFR in SuspecTed Coronary Artery Disease SUMMARY DEEPVESSEL FFR is a medical device that is designed to extract three- dimensional coronary tree structures and generate computed tomography-derived...
Multi-national multi-center clinical validation study ADAPT [1]
- DVFFR analysis was conducted on a total of 269 patients with 358 target vessels from 10 clinical sites (5 from US and 5 from EU)
- Primary endpoint: per-vessel sensitivity and specificity of DVFFR to detect ischemic condition comparing with invasive FFR measurement
- Study demonstrated that DVFFR yielded good diagnostic performance and met pre-specified criteria for study success
PER-VESSEL | Estimate, % | 2-Sided 95% CI (lower bound) | Target Rate | Met/Not Met |
Sensitivity | 86.9% | 80.6% | 75% | Met |
Specificity | 86.7% | 82.0% | 70% | Met |
PER-VESSEL | Estimate, % (95% CI) |
Accuracy | 86.8% (83.0-90.4%) |
PPV | 79.4% (71.8-86.2%) |
NPV | 91.9% (87.7-95.6%) |
PER-PATIENT | Estimate, % (95% CI) |
Sensitivity | 87.4% (79.4-93.1%) |
Specificity | 83.7% (76.5-89.4%) |
Accuracy | 85.2% (80.2-89.4%) |
PPV | 79.6% (71.0-86.6%) |
NPV | 90.1% (83.6-94.6%) |
Clinical Trials Performance Assessment
2017 Clinical Trial
In 2017, DeepVessel FFR was tested via a multi-center prospective clinical trial in which patients were recruited from:
- Beijing Anzhen Hospital affiliated to Capital Medical University
- Air Force General Hospital PLA
- Cangzhou Central Hospital
Study results demonstrated: | |
Accuracy | 89.2% |
Sensitivity | 94.2% |
Specificity | 84.3% |
Positive Predictive Value | 86.9% |
Negative Predictive Value | 92.2% |
2019 Clinical Trial
In 2019, DeepVessel FFR was tested via a multi-center retrospective clinical trial in which patients were recruited from:
- Beijing Anzhen Hospital affiliated with Capital Medical University
- Fuwai Hospital, Chinese Academy of Medical Sciences
- Beijing Friendship Hospital
- The First Affiliated Hospital of Wenzhou Medical University
- The Second Affiliated Hospital of Zhejiang University Medical College
- Wuhan Asia Heart Hospital
Study results demonstrated: | |
Accuracy | 90.8% |
Sensitivity | 94.7% |
Specificity | 88.6% |
Positive Predictive Value | 83.0% |
Negative Predictive Value | 96.7% |
[1] Wang ZQ, Zhou YJ, Zhao YX, et al. Diagnostic accuracy of a deep learning approach to calculate FFR from coronary CT angiography. J Geriatr Cardiol 2019;16:42-48.
Learn more about DEEPVESSEL FFR
We are actively looking for clinical partners in the United States and European Union.
