Dandelion Health and InVision Medical Technology are partnering to give life sciences companies up to 30x more cardiac measurements per patient, generated by running AI on raw echocardiogram data and layered onto a decade of longitudinal real-world patient records. The result: more precise phenotyping, measurable structural treatment effects, and trial designs built on cardiac physiology rather than billing codes.
NEW YORK, April 30, 2026 /PRNewswire-PRWeb/ -- AI generates cardiac measurements that never existed in the medical record, producing up to 30x more data points per patient from Dandelion's decade of real-world echo data
Dandelion Health, a real-world data and AI platform, today announced a partnership with InVision Medical Technology to bring InVision's unique echocardiography AI onto Dandelion's Clinical AI Marketplace. InVision's technology analyzes the raw imaging data captured during a cardiac ultrasound exam and generates automated interpretations, quantitative cardiac measurements, and disease risk assessments that were never recorded during clinical care. The integration gives life sciences companies a new class of cardiac data, including more precise assessments not obtained in usual clinical care, layered on top of Dandelion's longitudinal real-world dataset spanning over a decade of patient records, to power research across heart failure, ASCVD, cardiomyopathies, cardio-oncology, and the broader cardiometabolic landscape.
Echocardiograms are performed for a wide range of clinical indications, and physicians measure what is relevant to the patient in front of them — not the full set of measurements and interpretations a researcher might need. The raw imaging remains in internal data servers but are never again accessed. Dandelion's Clinical AI Marketplace changes this by running validated AI algorithms on the underlying imaging data to produce measurements that never existed before, and appending them directly to patient-level records to create research-ready datasets. This can lead to up to 30 times more data points being created.
"When we image the body, we capture a remarkable amount of data. But medicine has traditionally extracted only what it needs for the immediate diagnosis - the rest often goes unmeasured, which is a missed opportunity to understand so much more than we do today." said Elliott Green, Co-founder and CEO of Dandelion Health. "With InVision's technology now live on our Clinical AI Marketplace, we can generate those insights at scale and pair them with a decade of longitudinal patient data. That fundamentally improves how precisely you can measure treatment effects and use it to improve understanding of disease trajectories and design better clinical trials."
Why It Matters for Life Sciences
InVision's technology reads the raw imaging data from each echocardiogram and extracts nearly 50 measurements and 100 interpretations that reveal the structure and function of the heart: how large the chambers are, how well the walls contract, whether heart valves are stiff or leaky, how blood flows through the valves, and how the heart performs under load. These are the measurements that define whether a patient has heart failure and what kind, whether a therapy is changing the heart at the structural level, and whether a trial population is truly the one you intend to study.
The technology also brings the consistency needed for accurate research to these measurements. Physicians use different methods to measure the same parameter (even within the same health system) because there is no universal standard dictating which one to use. The resulting variability can obfuscate true insights and undermine trial design, making it harder to set reliable eligibility thresholds, compare outcomes across sites, or detect real treatment effects in the data. InVision's AI standardizes these measurements by applying the same American Society of Echocardiography recommended methodology to every echocardiogram, regardless of where or when it was performed.
"While the most common form of cardiac imaging, the Achilles' heel of echocardiography is variation in interpretation and measurement. Imprecision and inconsistency in interpretation masks clinically meaningful change, subtle disease progression, or early toxicity important to recognize in a disease program," said David Ouyang, Co-founder and CEO of InVision Medical Technology. " Our echo AI algorithms, trained on some of the largest echocardiography datasets in the world and validated in papers published in Nature, Nature Medicine, JAMA Cardiology, Circulation, NEJM AI, and other top venues, represent some of the most well-vetted and rigorously tested echo AI algorithms. Pairing this with the rich data in Dandelion Health's platform enables transformative understanding of real-world disease and treatment effects."
Validated for Research-Grade Confidence
Dandelion offers fit-for-purpose validation for algorithms on its Clinical AI Marketplace, confirming that an algorithm performs at a level sufficient for the specific research questions life sciences companies will use it to answer, and that it does so equitably across diverse patient populations. For InVision's echocardiography AI, Dandelion validated performance across the echo measurements most relevant for life sciences research, focusing on the subset most impactful for pharmaceutical use cases.
Algorithm outputs were compared against the same measures as reported by the original interpreting physician, drawn from both structured data and information abstracted from clinical notes. The echocardiograms used in the validation were sampled from a representative cohort of HFpEF patients, selected to reflect diversity across multiple health systems, BMI strata, age groups, and race/ethnicity — ensuring the validation captures real-world variability rather than performance under idealized conditions.
A New Data Layer for Cardiometabolic Drug Development
The partnership between Dandelion and InVision arrives at a moment of intense therapeutic innovation across cardiometabolic disease. More than a dozen therapies are in active development for HFpEF alone, alongside novel agents for ATTR cardiomyopathy, new ASCVD approaches, and GLP-1 receptor agonists with emerging cardiovascular benefits.
For the companies developing these therapies, the ability to phenotype patients using quantitative cardiac measurements, measure treatment effects at the structural level of the heart, and design trials around real-world imaging data (rather than billing codes) represents a fundamental shift in what's possible.
For life sciences companies looking to move faster, design smarter, and see deeper into their patient populations, the insights to do so are now available at scale.
About Dandelion Health
Dandelion Health is a real-world data and clinical AI platform powering next-generation precision medicine. Dandelion's platform delivers novel clinical insights at scale from rich, multimodal patient data, enabling AI, medical device, and life sciences companies to accelerate the entire product development lifecycle. Dandelion's Clinical AI Marketplace runs validated AI algorithms on multimodal real-world data to turn raw imaging and waveforms into structured, research-ready data for life sciences research and drug development. Dandelion is a growing consortium of leading healthcare systems from across the United States. To learn more, visit www.dandelionhealth.ai.
About InVision Medical Technology
InVision Medical Technology Corporation is an AI technology company improving the precision and accuracy of cardiovascular imaging. InVision's software tools improve the performance of echocardiography, the most common and easily accessible cardiac imaging modality. Backed by YCombinator, InVision is focused on serving patients and providers through the process of development and commercialization of AI diagnostics. Please visit invisionmedtech.com and follow on Twitter @InVision_AI.
Media Contact
AV Ploumpis, Dandelion Health, 1 201-472-0271, [email protected], https://dandelionhealth.ai
SOURCE Dandelion Health
Share this article