IMPRESS

The correct interpretation of electrocardiograms (ECGs) is crucial for accurately diagnosing cardiac abnormalities. Current methods, both manual by physicians and computerized, have not achieved the level of accuracy comparable to cardiologists in detecting acute cardiac issues. Leveraging advancements in artificial intelligence and big data, particularly deep neural networks, offers promising avenues to improve ECG interpretation where traditional methods have fallen short. The ECG-Project develops deep learning algorithms to automate ECG interpretation, particularly focusing on areas where current methods are inadequate. Through this research, we aim to revolutionize ECG interpretation, improving diagnostic accuracy, reducing healthcare resource utilization, and ultimately enhancing patient outcomes. The Research The project objectives are: WP1: Creating an algorithm capable of accurately and swiftly triaging ECGs through transfer learning, uncovering features in diseases with unknown ECG characteristics (such as primary arrhythmia syndromes and genetic disorders). WP2: design a portable multi-lead ecg-device, suitable for use by patients at home and healthcare professionals. This device will enable high-quality ECG acquisitions for rapid diagnosis. Origin This project is funded within the Innovative Medical Devices Initiative (IMDI) program 'Heart for Sustainable Care'. The focus of this program is the development of medical technology for the earlier detection, monitoring, and better treatment of cardiovascular diseases to ensure accessible healthcare and sufficient staffing. The program has been developed and funded by the Dutch Heart Foundation, ZonMw and NWO, who collaborate within the Dutch CardioVascular Alliance.

Until recently the standard approach of patients with recent-onset atrial fibrillation (AF) involved early cardioversion. In the latest ESC AF guidelines, a delayed cardioversion approach within 48 hours has been added to the recommendations. However, given the self-terminating and recurrent nature of AF, cardioversion may not always be necessary, and rate control medication could suffice to manage symptoms until spontaneous conversion to sinus rhythm occurs. The Research Continuous heart rhythm monitoring elucidated the recurrent and transient nature of recent-onset atrial fibrillation (AF). In the RACE7 we showed that a wait-and-see approach (WAS) in patients with recent-onset AF (rate control for symptom relief followed by delayed cardioversion if needed <48h) allows spontaneous conversion to sinus rhythm in 69% of patients, obviating active cardioversion. Recurrences within one month were seen in 30% of patients in both groups, i.e. the initially chosen strategy did not affect the recurrence pattern. Considering the latter, it remains unclear whether cardioversion is needed at all, especially since cardioversion strategy does not seem to affect behaviour of the arrhythmia over time. Instead of cardioversion a watchful-waiting rate control strategy may be appropriate as initial strategy. Therefore, we intend to perform a multi-center clinical randomized controlled trial to show non-inferiority of watchful-waiting with rate control versus the WAS approach in terms of prevalence of sinus rhythm at 4 weeks follow-up, using a novel telemonitoring infrastructure to guide rate and rhythm control during follow-up. This novel telemonitoring infrastructure may facilitate the watchful-waiting strategy and obviate the need for cardioversion and reduce costs compared to the delayed rhythm control WAS strategy. The study will be conducted across multiple centers in the Netherlands, including UMC Groningen, Radboud UMC, Amsterdam UMC, Alrijne Hospital, VieCuri Medical Centre, Zuyderland Medical Centre, Elisabeth-TweeSteden Hospital, Rijnstate Hospital, Martini Hospital, St. Antonius Hospital, Antonius Hospital, Noordwest Hospitalgroup, Medisch Spectrum Twente, and Maastricht University Medical Center. Origin This project is funded within the Innovative Medical Devices Initiative (IMDI) program 'Heart for Sustainable Care'. The focus of this program is the development of medical technology for the earlier detection, monitoring, and better treatment of cardiovascular diseases to ensure accessible healthcare and sufficient staffing. The program has been developed en funded by the Dutch Heart Foundation, ZonMw and NWO, who collaborate within the Dutch CardioVascular Alliance.