In the field of cardiac healthcare, our continuous collaboration with Abbott stands as a beacon of innovation. Within this partnership, we present two pivotal publications that underscore the transformative impact of Idoven's artificial intelligence (AI) platform, Willem, on Abbott's ICM Confirm RxTM device for arrhythmia detection.

In our first breakthrough featured in the August 2022 issue of the Cardiovascular Digital Health Journal, Willem's integration with Abbott's ICM device achieved an impressive 95.4% accuracy for arrhythmia classification, coupled with a remarkable 98% reduction in false positives.

Building on this success, our second collaborative paper, recently published in the Journal of Electrocardiology, aimed to showcase the capability of Idoven's AI platform in expanding ICM arrhythmia recognition beyond the initial four patterns detected by Abbott's ICM device.

Our research underscores AI's ability to swiftly recognize a diverse range of cardiac patterns, elevating diagnostic capabilities beyond conventional standards. By efficiently handling complex data, AI enables faster interventions and personalized treatment, complementing our expertise for a more comprehensive understanding of heart health.‍

‍- Fabio Quarteri, Cardiologist at Ospedale S. Maria Nuova, Reggio Emilia.

What are the key outcomes of this collaboration?

The research, which is led by Abbott, the Hospital of Reggio Emilia, and Idoven and in collaboration with the Centro Nacional de Investigaciones Cardiovasculares (CNIC), Centro Español de Red de Investigación Biomédica en Enfermedades Cardiovasculares (CIBERCV), and Interhospital Foundation for Cardiovascular Research (FIC), has demonstrated that the use of AI significantly amplifies the detectability of cardiac patterns by Implantable Cardiac Monitors (ICMs), leading to a substantial improvement in the diagnosis of cardiovascular diseases.

Overall, this paper shows that using Willem to analyze Abbott’s ICM devices enables:

• Expanded detection from 4 to 25 cardiac patterns.
• Achieved an average accuracy of 88% in classifying 25 different cardiac rhythm patterns.
• Significantly reduced interpretation time, from 11 minutes to 6 seconds.

‍How was the study conducted?

This retrospective analysis involved a thorough examination of ECG data from 20 patients who were equipped with Abbott’s Confirm Rx™ ICM. The ECG data were recorded and analyzed by Idoven’s AI, Willem™, and cardiologists simultaneously.

The integration process unfolded as follows: first, all ICM ECG recordings were uploaded to Abbott’s cloud database, Merlin.net. Then, the recordings were converted to a digital format and transmitted to the Willem cloud-based AI platform for interpretation and processing, resulting in the generation of results and reports classifying cardiac episodes.

Following the study, it was determined that among nineteen patients, ICMs recorded 2261 subcutaneous ECGs (sECGs) over an average follow-up of 23 months. Within these recorded episodes, Idoven’s AI platform successfully identified and classified 7882 events, attributing them to 25 different cardiac rhythm patterns. The overall accuracy, factoring in the weighted contributions of each pattern, reached  88%.

In summary, the study confirms that Idoven's AI platform, Willem, enhances cardiologists' diagnosis and efficiency in processing ECG data from Abbott's ConfirmRx ICM devices. Its capability to detect various cardiac rhythm patterns, including rare disorders, highlights the potential of Idoven's AI to significantly assist healthcare professionals, reducing their workload, enabling more precise diagnoses, and ultimately optimizing patient treatments.

To gain a more in-depth understanding of this study, access the full paper here: https://www.sciencedirect.com/science/article/abs/pii/S0022073623001723