The Center for Data Innovation recently spoke with Onni Eriksson, cofounder of AIATELLA, a Finnish company using AI-powered software to analyze cardiovascular imaging more efficiently. Eriksson discussed how the company trains its algorithms on diverse datasets to reflect a wide range of patient demographics and heart conditions, ensuring that the results are both accurate and transparent for medical professionals.
David Kertai: How does AIATELLA use AI to turn data into useful insights?
Onni Eriksson: Our AI platform uses advanced computer vision and deep learning algorithms to analyze medical images, like CT scans, and extract precise cardiovascular measurements that would typically require extensive manual work. The system automatically detects important structures in the heart and blood vessels and measures things like vessel diameter, wall thickness, and blood flow characteristics. A human specialist would typically spend 30 to 45 minutes doing this by hand, but our software completes it in just a few minutes. This shift allows doctors to focus more on patient care and clinical decision-making rather than manual measurements.
It doesn’t just automate routine tasks, it also flags potential problems, monitors how conditions change over time, and generates reports. By turning complex image data into structured, easy-to-understand insights, we enable faster diagnoses and better treatment decisions, ultimately leading to improved outcomes.
Kertai: How do you make sure your technology works in different medical settings?
Eriksson: We partner with hospitals, imaging centers, and medical professionals across multiple continents and test our solutions with varied imaging equipment, protocols, and patient populations. This approach helps us address differences in imaging quality, technical standards, and clinical practices that exist between regions and healthcare systems.
Our platform is hardware-agnostic, meaning it works with images from a variety of machines, whether they’re in cutting-edge hospitals or clinics with limited resources. We’ve tested it with patients of different ages, body types, and cardiovascular conditions to ensure consistent performance. Additionally, we stay in close contact with the doctors who use our system, so we can continuously improve the algorithms and user experience based on real-world feedback.
Kertai: How do you keep your AI system accurate and trustworthy?
Eriksson: Trust is essential when it comes to AI in medicine. We’ve built our system around three key principles:
First, we train our AI system on data from a wide range of patients with different backgrounds, conditions, and body types. This helps us reduce bias and ensures the system works well for everyone, not just a narrow group.
Second, we focus on explainable AI. Our system doesn’t just give you a result, it shows exactly what it measured and how. For example, if it calculates the width of a blood vessel, a doctor can see where and how the measurement was taken. If needed, they can adjust or override it.
Third, we maintain ongoing quality checks. We compare the system’s results with expert annotations, conduct regular performance reviews, and collect feedback from users. The system also provides confidence scores that alerts doctors when image quality or other factors might reduce accuracy. These steps ensure the technology remains transparent and reliable.
Kertai: Can you share an example of how AIATELLA has helped in heart care?
Eriksson: One of our most impactful experiences happened during a medical conference where we offered complimentary carotid artery screenings to demonstrate our technology. What started as a product demonstration turned into a potentially life-saving intervention for several attendees. Our AI system identified multiple individuals with significant carotid stenosis, dangerous narrowing of the arteries that supply blood to the brain, that had gone completely undiagnosed. These people weren’t feeling unwell, they were walking around the conference floor, completely unaware they were at high risk for stroke. We immediately connected them with vascular specialists for follow-up care and interventions.
That moment reinforced why we do this work. It showed the enormous value of quick, accessible heart screenings. Millions of people are living with silent but serious heart and vascular problems. If we can catch these early, before symptoms appear, we can prevent countless strokes, heart attacks, and deaths. That’s the future we’re working toward.
Kertai: What’s next for AIATELLA?
Eriksson: We’re expanding our impact across the entire spectrum of heart and blood vessel care. Right now, we specialize in CT imaging, but we’re also moving into other types of imaging like MRI and ultrasound. These tools help assess different parts of the body, including the carotid arteries in the neck and the peripheral vascular system, which includes blood vessels in the arms and legs.
We’re also building tools to help doctors track how cardiovascular disease progresses over time. That could help them make better decisions about when to intervene. On a broader level, we’re exploring how our technology can support population health efforts, for example, making large-scale heart screenings possible in places where resources are limited. Our vision is to make advanced cardiovascular screening as routine and accessible as checking your blood pressure.