CNN recently featured our partnership with Abbott, a global leader in medical devices, showcasing the work we are doing together in hospitals across Finland in its program Tomorrow Transformed. By integrating Willem™, Idoven’s AI technology, into the data triage workflow of Abbott’s implantable cardiac monitors (ICMs), this collaboration is setting new standards in cardiac care.
Driving innovation in Finnish Hospitals
This collaboration began at Tampere University Hospital (TAYS), with our AI processing cardiac events from Abbott’s implantable cardiac monitors (ICMs). This integration enables cardiologists to filter “events” (irregular heart activities detected by the device) for review, reducing their workload and allowing them to focus on urgent cases, thereby improving patient care.
We have since expanded this project to other hospitals across Finland, including HUS Helsinki University Hospital, Hospital Nova, and Hämeenlinna Heart Hospital. In addition to continuing this support for cardiologists, we are beginning to extend our technology to other specialties as well. For example, at HUS, neurologists —often without direct cardiology expertise— are now relying on our technology to review and classify episodes from Abbott’s ICMs, particularly in identifying cardiac causes of strokes or syncope.
How it all began: From research to real-world impact
Our collaboration with Abbott started several years ago with a shared mission: to leverage AI for the early detection of cardiovascular diseases. Abbott’s ICMs generate vast amounts of data by continuously monitoring heart rhythms, making manual reviews time-consuming. We saw the potential of combining Abbott’s innovative devices with Idoven’s AI to improve arrhythmia detection.
In the research phase, we worked with Reggio Emilia Hospital in Italy to validate our AI’s effectiveness in cardiac diagnostics. The results were game-changing:
- 98% reduction in false positives, significantly decreasing unnecessary reviews by healthcare professionals.
- 95.6% accuracy in detecting arrhythmias, approaching cardiologist-level precision.
- Expansion from detecting 4 to 25 cardiac patterns, offering a much more detailed analysis.
- Reduced analysis time from 11 minutes to just 6 seconds, speeding up data interpretation.
These findings were published in two key journals: "Artificial Intelligence Augments Detection Accuracy of Cardiac ICMs" (2022) in the Cardiovascular Digital Health Journal and "AI Cloud Platform Improves Arrhythmia Detection from ICMs to 25 Cardiac Rhythm Patterns through Multi-label Classification" (2023) in the Journal of Electrocardiology.
Watch the full episode of CNN’s Tomorrow Transformed to see how AI and digital technologies are transforming healthcare here.