Ultromics | Ground-breaking cloud AI automation for echo analysis

Ultromics cloud service, EchoGo, is an automation tool that leverages the world’s most sophisticated artificial intelligence and the Microsoft Azure cloud to provide high quality, rapid interpretation of echocardiograms for earlier heart disease detection.

CE-marked and FDA-cleared, the patented technology is used worldwide by over 50 sites – including the NHS, Mayo Clinic and MedStar - and is helpful for the most common markers of various cardiac diseases, including left ventricular ejection fraction and cardiac strain.

The technology is supported the NHSX, as one of their grant winners to roll out the technology into the NHS this year.

Seamless automation that improves and standardises care

There are around 1.2 million echo scans performed in the UK. Current software for interpreting echocardiogram scans requires partial manual tracing of endocardial borders which is time consuming and operator dependent.

Furthermore, teams have traditionally used EF to measure left ventricular systolic function, which is a marker of HF. But it is imperfect. It can be subjective, making it unreliable, and changes can often trail behind disease progression. More advanced detection and monitoring methods are needed, and there is growing evidence that global longitudinal strain (GLS) is more useful than EF in this space, but it has yet to move into the mainstream of echo assessment.

In fact, fewer than 10% of NHS Trusts routinely measure GLS, despite the fact more than half of UK heart failure patients, or 1.3 million people, have HF. Of these, 50% have preserved ejection fraction (HFpEF).

EchoGo’s fully automated service uses AI and machine learning to provide rapid, reproducible measurements of left ventricular volumes, EF and cardiac strain - with the same precise results every time, all the time.

Practices can take advantage of EchoGo automating both EF and Strain, to better detect heart failure and pick up patients with preserved ejection fraction.

Results show that EchoGo is so precise and accurate that it can outperform manual analysis when predicting clinical patient outcomes. With lower variability and greater review efficiency, clinicians can more accurately determine a person’s likelihood of developing heart issues that require management or treatment.

Cloud automation can have a greater impact and ROI for organisations, and can provide improved productivity and security. EchoGo updates instantly and automatically so sites always have the latest tech and science innovation.

Real world implementation

EchoGo has been validated across 50+ sites worldwide, including 30 NHS Trusts in the UK, and many sites in the US including Mayo Clinic and MedStar.

Because of the way EchoGo is connected through a single seamless cloud connection, EchoGo can be set up and live within 48-hours after contracting, with flexible pay-per-scan tiers that allow you to scale up easily as use cases grow.

EchoGo operates through the Microsoft Azure cloud and connects directly into a customer’s PACS. The software uses machine learning algorithms to run behind the scenes to automatically analyse echocardiograms, imitating the way the human eye identifies borders and measurements. EchoGo automatically selects the best view, contours and sends a report back to site to support clinical decision-making, with reports accessible from anywhere.

The technology meets the industry’s highest data security standards (ISO 27001) and is compliments with Data Protection Regulation (GDPR) and the U.S. Health Insurance Portability and Accountability Act (HIPAA).

AI you can trust for your patients, with proven results

The foundations of EchoGo’s ground-breaking technology and patented Artificial Intelligence has been trained and validated on over 50 million data points. Its AI algorithm was trained from a one-of-kind outcomes based dataset ethically sourced from around the world, enabling Ultromics to not only provide high-fidelity diagnostic and prognostic indicators but also accurately predict outcomes such as mortality and extended hospitalisations.

Spun out of the University of Oxford, one of the world’s most prestigious scientific and academic venues, EchoGo was built with the leading minds in the space of Artificial Intelligence and Machine Learning, and has been refined and validated over many years using the most rigorous scientific best practices, with leadings organisations.

EchoGo has been tested and validated to provide zero variability between operators. Accredited experts compared EchoGo with a predicate AI device using manual interaction. Hundreds of echo studies were processed and compared. EchoGo produced 0% variability in measuring left ventricular ejection fraction, LV volumes and GLS between operators, providing better precision and excellent reliability. Experts using the traditional predicate software produced a high level of inter- and intra- operator variability.

Stop by the Ultromics booth at BSEcho 2021 to learn more about the benefits of EchoGo and get a free demo.

To learn more about EchoGo visit Ultromics’ website or follow us on Twitter and LinkedIn.