AI Helps Save Hearts

Cardiovascular disease kills 17 million people each year. Idoven’s founders, Dr. Manuel Marina Breysse and José María Lillo Castellano, together with their team, created Willem, an AI that reads electrocardiograms (ECGs) with expert-level accuracy. Their goal: to provide faster diagnoses for doctors, enhance patient care and foster a future in which AI transforms cardiology.

How did you get the idea to use AI to fight heart disease?

Manuel: I’m a cardiologist. Every day at the hospital I saw patients suffering from cardiovascular disease, the leading cause of death worldwide. One in three people will die from this disease unless we improve our healthcare systems. It’s also the biggest cost driver in healthcare. Diagnosing it starts with reading ECGs, a repetitive and time-consuming task. José and I were convinced there was a better way: to build an AI that can read ECGs.

When was that and what happened in the preparation phase
before the foundation of your startup Idoven?

José: We founded Idoven in 2018, building on our collaborative research and PhD work in machine learning applied to cardiology which began in 2014. Willem is our AI platform, capable of detecting arrhythmias and diseases in ECGs.

Can you describe your idea in more detail?

Manuel: At Idoven, we are creating Willem, an AI cardiologist expert in diagnosing ECGs of any duration. It is already possible to use Willem as a copilot agent, accessible through APIs, to build a wide variety of cardiovascular solutions.

Why is this important?

José: Today, healthcare professionals around the world spend millions of hours analysing ECGs daily. A piece of software can do that in milliseconds. So, the goal is not only to improve diagnostic accuracy, but also to free up doctors so that they can focus on patient care.

What does that mean for patients?

Manuel: Access to healthcare professionals and time are two of the biggest barriers. It can take months for a patient to get a referral to a cardiologist. AI can reduce those waiting times and help diagnose patients faster. That changes how the disease is managed. Earlier diagnosis normally leads to better outcomes at lower cost. Patients suffer less, and mortality rates can be reduced.

Who uses Idoven technology?

Manuel: Our clients are pharma, medtech companies, healthcare systems and global organisations. Our AI platform enables researchers, developers, healthcare leaders and organisations to achieve breakthroughs and solve various problems where an ECG is involved. The end user is a doctor, and both patients and doctors benefit from our AI. Over 566,900 patients and doctors have trusted us and collaborated with our team of engineers, scientists and healthcare professionals to train Willem with their ECGs. They did so through the #DonateYourHeartbeats to save lives movement. Created by Idoven and supported in 2019 by Andrus Ansip, then Vice President of the European Commission, the initiative has received strong support from citizens, patients and over 150 leading research hospitals and organisations in 33 countries.

You started partnering with Google in 2019. Why did you do that?

José: Starting a company comes with uncertainty, and you often feel alone. Seven years ago, Manuel knocked on Google’s door as we were looking for a workspace. That’s where it all started. We applied to the Residency programme and became alumni of Google for Startups. It was one of the best decisions we made. At the time, we were just three people. Now we’re a team of 58.

What does the partnership entail?

José: The Google for Startups programme allowed us to connect with founders and startups. This was one of the most valuable aspects. In 2019, we received a grant that was reserved at that time for US universities: €100,000 in Google Cloud credits, which gave us access to powerful computing and tools to validate key research ideas. It was a real turning point. We also had the chance to work with and learn from Google experts from Silicon Valley who came to Madrid, some of whom are still friends and advisers.

How important are partnerships for you in general, e.g. with pharmaceutical companies?

Manuel: Our mission is to serve humanity by creating an advanced AI platform for ECGs that helps organisations transform diagnostic pathways. Partnerships are essential: we collaborate with global pharma teams, medtech and tech leaders, leading physicians and some of the most innovative hospitals and research centres. Our AI platform empowers them to prevent life-threatening conditions and bring smarter diagnostic tools for faster diagnoses and better treatment decisions.

According to a Google for Startups Report, AI adoption among European startups remains low. How important are companies like yours for Europe’s competitiveness in AI and the economy?

Manuel: I have worked in hospitals, research centres and public organisations. Now, working in a startup, I see them as crucial structures for rapid, product-oriented innovation. The beauty of great products is that they can solve a huge problem at scale, benefiting both patients and society. I see deeptech AI startups like ours as key organisations that society should support, learn how to collaborate with and protect. The ones that succeed are rare, but those that do are vital to shaping the future of Europe and the world, as has happened many times.

What do you think about AI’s potential in medicine in general? Will it help humankind to defeat all manner of serious diseases in the not-so-distant future?

Manuel: AI is changing the world very quickly. It’s reducing the burden of repetitive tasks, the kind of routine work done every day in hospitals, in our case with near-cardiologist accuracy. That allows doctors and healthcare experts to rethink how medicine will work in the years to come. For decades, innovation in medicine came through hardware or new treatments. Today, it lies in AI, in our ability to build software tools that let us measure what we couldn’t before. And when we can measure, we can understand, improve and create knowledge. By being accessible and useful, this knowledge moves medicine forward for generations to come.

What’s your vision for what Idoven will achieve in the next ten years?

Manuel: Our vision is to build a foundational AI model for signal processing. We’re creating something both special and challenging, because the problem we’re tackling is urgent. There’s only one cardiologist for every 32,000 people in the world, and even after more than a decade of medical education, much of their work involves repetitive tasks that could be easily handled by AI. We want to change that by giving doctors tools that let them focus on complex care. To achieve this, we’re partnering with leading healthcare and research institutions who share our belief in the potential of this technology to help patients everywhere.