Research

Clinical Validation

Our platform is backed by peer-reviewed research demonstrating clinical efficacy and improved outcomes.

CathEF Study: AI-Powered Cardiac Function Analysis

Robert Avram et al.

JAMA Cardiology, 2023

This groundbreaking study demonstrates the effectiveness of AI-powered analysis in cardiac function assessment, achieving remarkable accuracy while significantly reducing analysis time.

0.90 AUC with TTE LVEF measurements

4-second analysis time per case

Novel biomarker extraction techniques

Validated against clinical standards

DeepRV: Automated Right Ventricular Function from Coronary Angiograms

Fatima Zahra Fawzi et al.

European Heart Journal – Digital Health, 2026

A video-based deep learning model that assesses right ventricular systolic function directly from routine coronary angiograms, removing the need for prior imaging. Developed at the Montreal Heart Institute, externally validated at UCSF, and deployed prospectively during primary PCI for STEMI.

0.80 AUROC internally, 0.75 on external validation

0.83 AUROC in prospective STEMI deployment

5.1-second median inference time

Raised cardiologist accuracy from 72.1% to 77.6%

DeepCORO-CLIP: A Multi-View Foundation Model for Coronary Angiography

Sarra Harrabi et al.

arXiv preprint, 2026

A multi-view foundation model trained with video-text contrastive learning for comprehensive coronary angiography interpretation, with external validation and publicly released code and weights.

Trained on over 203,000 angiography videos

0.888 AUROC internally, 0.89 on external validation

Detects chronic total occlusion, thrombus and calcification

4.2-second mean inference; open code and weights

EchoPrime: A Multi-View Vision-Language Model for Echocardiography

Milos Vukadinovic et al.

Nature, 2026

A view-informed vision-language foundation model for echocardiography that interprets every video in a complete study rather than a single view, using contrastive learning and retrieval-augmented interpretation.

Trained on 12 million video-report pairs

State of the art across 23 benchmarks

View-informed anatomic attention across standard views

Interprets the complete study, not a single video

PanEcho: Complete AI-Enabled Echocardiography Interpretation

Gregory Holste et al.

JAMA, 2025

A view-agnostic multi-task model that performs the full range of echocardiographic reporting tasks from a single pass, validated internally and across four external cohorts, and publicly released.

39 interpretation tasks from one model

0.91 median AUC across 18 classification tasks

LVEF estimated within 4.2% internally, 4.5% externally

Trained on over one million echocardiogram videos

CardioSyntax: End-to-End SYNTAX Score Prediction from Angiography

Alexander Ponomarchuk et al.

arXiv preprint, 2024

An automatic method for estimating the SYNTAX score of coronary disease severity from multi-view angiography video, published together with a purpose-built dataset and benchmark.

CardioSYNTAX dataset of 3,018 patients released

0.51 coefficient of determination for score prediction

77.3% accuracy for zero-score classification

Multi-view X-ray video of complete angiography