Book a Demo
Platform Overview
Features
Cardiac Amyloidosis Heart Failure Aortic Stenosis Pulmonary Hypertension Strain Stress Echocardiography
Operations
Operational Impact Deployment & Security Integrations Core Lab Solutions EHR Mining Partnerships
Tools
ROI Calculator Reimbursements Peer Comparison Regulatory Status
Library & news
Clinical Evidence Case Studies News
Our Story Careers Media Kit

AI echo analysis at Brigham's Cardiac Imaging Core Laboratory

The Cardiac Imaging Core Laboratory at Brigham and Women's Hospital deploys Us2.ai to speed echocardiographic analysis for clinical trials of novel cardiovascular therapies, combining automated reads with expert clinician review.

Brigham and Women's Hospital Division of Cardiovascular Medicine Boston, USA

At a glance

< 2 min analysis time per study (vs ~30 min manual)
0 variability between runs on the same study
2022 Nature Communications validation paper
2023 partnership announced

Echo analysis at clinical-trial scale

Heart disease remains the leading cause of death worldwide. Echocardiography is the affordable, front-line tool used to diagnose and assess it, but acquiring, measuring, and analysing the images is time-consuming. For an academic core laboratory running echocardiography for multi-site cardiovascular trials, every additional minute of analysis time per study multiplies across thousands of studies and slows the timeline for getting novel therapies to patients.

The Cardiac Imaging Core Laboratory at Brigham

The Cardiac Imaging Core Laboratory (CICL) is part of the Division of Cardiovascular Medicine at Brigham and Women's Hospital. Directed by Dr. Scott Solomon, Professor of Medicine at Harvard Medical School, the CICL provides centralized echocardiographic analysis for cardiovascular clinical trials. Us2.ai provides AI software that assists in analysing the research echocardiograms processed through the laboratory, with the goal of improving the speed and scalability of the CICL's read pipeline.

AI-assisted read, expert review

After an echocardiographer acquires a study, the FDA and CE cleared Us2.ai software automatically analyses every chamber using both 2D and Doppler views, producing a full structured report covering cardiac volumes (all four chambers), M-mode measurements including tricuspid annular plane systolic excursion, spectral Doppler across all valves (both PW and CW), and tissue Doppler. The measurements cover the majority of standard adult transthoracic echocardiography parameters recommended by the American Society of Echocardiography, the European Association of Cardiovascular Imaging, and the British Society of Echocardiography.

Reported analysis time drops from roughly 30 minutes per study to under two minutes, with zero variability between runs on the same study and accuracy comparable to expert clinicians. The AI output feeds into the CICL's existing expert-review workflow rather than replacing it: clinicians retain oversight on each report.

Validated by the lab that uses it

AI-assisted echocardiography analysis allows for faster, more reproducible and accurate assessment of echocardiographic changes that occur in response to novel therapies being tested in clinical trial. We validated this in our own laboratory, and it has proven comparable to human sonographers for a vast number of echocardiographic measures. Combined with expert review and assessment of echocardiograms, this approach will allow us to get novel therapies to patients faster.

Dr. Scott Solomon, Director of the Cardiac Imaging Core Laboratory, Brigham and Women's Hospital; Professor of Medicine, Harvard Medical School

The validation referenced by Dr. Solomon was published in Nature Communications in 2022, showing that fully automated Us2.ai measurements were comparable to expert human measurements across a large set of echocardiographic parameters, and that the AI analysis was fully reproducible for a given patient study.

Faster trials, better evidence

Cardiovascular clinical trials sit on echocardiographic endpoints. Reducing the bottleneck in core-lab read times reduces the bottleneck in evidence generation, which is the bottleneck in getting novel therapies through development and to patients. For trial sponsors, the value is concrete: AI handles the measurement work end-to-end while the academic core lab focuses its expert time where it matters most.