Stargaze RisingStar Series: Zachi Attia
A pioneering researcher bridging the gap between AI tools and clinicians
This post is the fifth in a series on Stargaze™, Portal Innovations’ AI toolkit for sourcing, mapping, and predicting innovation. You can read the first four posts here:
A decade ago, Paul Friedman, a cardiology researcher at Mayo Clinic, went to Ben Gurion University in Tel Aviv to recruit an electrical engineer. His mission was to bring a unique perspective to a significant problem; there was a stockpile of cardiology data being underused and growing at a level that clinicians couldn’t keep up with. This mission strongly resonated with Zachi Attia and enticed him to become a visiting scholar at Mayo.
Around the same time, we began brainstorming the Stargaze platform to solve a similar problem. But instead of looking for people at risk of cardiac disease, our objective was to look for people “at risk” of creating biotech breakthroughs. And instead of using cardiology data, we used scientific metadata. Different data sets, different objectives, but the same underlying tools and technology.
Fast forward to 2024: Portal pointed Stargaze at the Mayo Clinic to tell the story of Mayo professor David Ahlquist and the re-founding of Exact Sciences - a molecular diagnostics company now worth over $10 billion - out of his technology. Our question: “What was it about Ahlquist’s work at Mayo that enabled him to create such a transformational technology? What can we learn from his story? Are there patterns in his career that could help us find other startups like Exact Sciences before they hit big?”
Using our 7-step Stargaze recipe for finding up-and-coming innovators, we immediately zoomed in on Dr. Attia, called him up, and asked him what he was working on.
Dr. Attia’s Story
Zachi Attia’s educational background is in electrical and software engineering, so cardiology was a foreign language to him at first. However, Dr. Friedman’s initial project was detecting potassium levels from ECG data, and data analysis was something Attia was already an expert in. This project quickly wrapped up and was licensed out, but it was enough to entice Attia to stay at Mayo Clinic as a full-time researcher. Attia was the first in a newly formed AI group within the Cardiology department, one of the first of its kind across the world. He started leveraging traditional ML and neural networks on ECG data to answer the Goliath of problems in cardiology - getting ejection fraction data from an ECG. From this clinicians can predict heart failure, cardiomyopathies, valve disorders, and more.
Zachi’s advanced data analytic platform was not only able to solve and capture ejection fraction levels from ECGs, but also in a manner that can be clearly communicated to clinicians at scale. However, solving a problem in a lab is not the same as solving it for patients. So Zachi sent his AI dashboard to some of his clinician friends. It went viral. Soon clinicians all over were using his dashboard to assist in diagnoses and data analyses in the cardiac space.
It was great to hear that clinicians were using it, but Attia wanted to make sure it was tangibly helping with patient outcomes. He designed a randomized clinical trial to evaluate two main questions: is his dashboard improving diagnostic accuracy and is it decreasing resources necessary to make decisions? The trial was an enormous success. It revealed that the clinicians using his AI dashboard found over 30% more cases of patients with low ejection fraction with no additional resources or testing required.
The Mayo Clinic is one of the hospitals at the forefront of integrating AI into their clinic, and Dr. Attia is a pioneer in this space. They are one of the first to create a new position, held by Attia and 4 others, as an “AI Consultant” in the clinic. Despite having no formal medical training (he is a PhD, not an MD), his role in the Department of Cardiovascular Medicine is to directly advise clinicians on matters related to patient care. In the clinic his role is as a translator, communicating AI outputs to clinicians to improve patient outcomes. Doctors are already asked to be experts in a wide breadth of subjects; most don’t want to add AI to that list. Instead, experts like Zachi are being asked to have a voice in the clinic, advising physicians on how to incorporate AI outputs into their diagnoses and treatment of patients.
A Stargaze RisingStar
Since being at Mayo Clinic full-time, Attia has assisted in founding and serves as an advisor for two startups - Anumana and Marani:
Anumana uses data from >11M patients, combined with machine learning and scientific expertise, to create AI solutions that seamlessly integrate into the clinic. They have an FDA-approved algorithm AI-ECG LEF to detect low left ventricular ejection fraction. Also, they are working on synthesizing insights from ECGs, EGMs, and EHRs to inform peri-procedural decision-making.
Marani is a prenatal and postpartum care company, developing AI-powered devices for fetal and maternal monitoring. Their goal is to provide remote monitoring for clinicians while giving patients direct insight into their own pregnancy.
Attia is also working on a third startup that is currently in stealth mode, so watch out for that on the horizon. As we’ve rolled out Stargaze to find RisingStars, we’ve seen this trend repeat over and over: the RisingStars we spot are almost always working on a startup.
It makes sense. If you train an algorithm to find the most promising, young biotech innovators, it is going to be exceptionally suited to finding stealth and pre-launch startups, too. If you’re looking for stealth, low-profile, or pre-incorporation startups led by the most talented young biotech researchers, Stargaze is the fastest way to find them. There are a lot more Zachi Attia’s out there.
Interested in exploring Stargaze yourself? Sign up to become a Stargaze Astronomer.
Join us for a virtual event, “Stargazing with Portal Innovations - Featuring Zhenpeng Qin” on March 22nd. We’ll be interviewing and further discussing the work of Zhenpeng Qin - another RisingStar identified by Portal Stargaze.