Eko raises $20 million for AI heart disease monitoring and detection solutions

Heart disease is a silent and often costly killer. About 610,000 people die of it […]

Heart disease is a silent and often costly killer. About 610,000 people die of it in the U.S. each year, putting it at the top of the leading causes of death for women and men. And the total combined annual cost in the U.S. arising from heart disease is roughly $200 billion, owing to health care services, medications, and lost productivity.
Since 2013, a trio of entrepreneurs — Connor Landgraf, Jason Bellet, and Tyler Crouch — has been attempting to turn the tide with heart disease monitoring software, AI, and heart sound and electrocardiogram (ECG) sensors. Their work catalyzed the founding of Berkeley, California-based Eko Health, which partnered with the Mayo Clinic on algorithms that screen patients for the presence of weak heart pumps. Separately, Eko is collaborating with Northwestern Medicine and Sutter Health to improve valvular heart disease screening and connect cardiologists with remote rural patients.
It’s in the midst of this flurry of activity that Eko secured $20 million from Artis Ventures, with contributions from a raft of existing backers including DigiTx Partners, NTT Venture Capital, 3M Ventures, the Mayo Clinic, Seraph Group, and XTX Ventures. The fresh capital brings Eko’s total raised to nearly $30 million, which CEO Landgraf said will be used to accelerate research and product development.
“With the staggering number of people affected by heart disease, it’s clear that advancements in cardiac screening and monitoring have never been more urgent,” he added. “Artificial intelligence is arguably one of the most powerful advancements in modern medicine, enabling clinicians to predict with more accuracy, diagnose with more confidence, and in the end, give their patients the best care possible.”
Eko’s suite of AI algorithms, which Landgraf carefully notes haven’t yet been cleared by the FDA, ingest heart sound and ECG data to spot structural heart disease, rhythmic abnormalities, atrial fibrillation, low ejection fraction, and other signs of heart failure. Their predictions are delivered to a HIPAA-compliant mobile app and web dashboard, from where clinicians can manage, share, and livestream them during remote telemedical visits. The raw metrics feed into Eko’s patient-facing Eko Home, which enables patients to send daily heart sound and ECG recording back to their care team.

As Landgraf explains, heart sounds typically consist of two regular sounds — S1 and S2 — for every isolated heartbeat. Normal blood flow inside the heart is silent, but disrupted or turbulent flow causes the surrounding tissue to vibrate and produce an audible murmur.

Murmurs overlap with heartbeats, making them difficult to perceive with a traditional stethoscope. But Eko claims its AI models trained on multiple data sets of heart sounds can accurately differentiate between healthy sounds and those containing murmurs. Specifically, they pin down the location of the S1 and S2 sounds to specify whether a given murmur is systolic or diastolic, and they’re able to suss out the presence of atrial fibrillation and normal sinus rhythm as well as chronic conditions like tachycardia (abnormally high heart rate), bradycardia (abnormally slow heart rate), and premature ventricular contractions. Additionally, they can measure electromechanical activation time (EMAT), a systolic interval that’s sometimes used to measure the electromechanical desynchronization of the heart.

Two key parts of Eko’s product ecosystem are the FDA-cleared Core and Duo stethoscopes, the former of which delivers a claimed 40 times sound amplification for off-the-shelf stethoscopes or as a standalone replacement. As for the Duo, the company asserts it’s the first stethoscope to combine heart, lung, and bowel sounds with ECG.
Eko says its products are used by “tens of thousands” of clinicians treating “millions” of patients at over 3,000 hospitals and providers around the world. By the end of the year, the company expects to begin offering drug-data combinations matched to recordings by Eko Home with other life science partners, and it plans to make available a software developer kit for hospitals and health care providers who wish to build Eko Home platform directly into their apps.
“Eko is transforming cardiac care as we know it,” said Artis Ventures partner Vasudev Bailey. “They are the perfect example of how machine learning using quality data sets can positively influence patient outcomes and improve quality of care. This demonstrates the potential for an immense pipeline of life-saving applications where sound can aid in the screening of many other diseases.”

Original Article: (http://feedproxy.google.com/~r/venturebeat/SZYF/~3/kVASLFSTTCc/)