Dealing with ECG data overload with Artificial Intelligence

An ECG is a recording of the heart's electrical activity. The amount of ECG data being collected has increased dramatically in recent years with the advent of more sophisticated ECG monitoring devices and latest-generation smartwatches, which are able to record ECGs at any time. This increase in data production leads to significant challenges for both healthcare providers and patients when managing this data overload.

An explosion of new devices

There were over 43 million Apple Watches sold in 2020, and this number is only going to increase in the coming years. The Apple Watch is now approved by the FDA as a cardiac monitor that can be used to automatically detect atrial fibrillation. With such a large installed base, it is inevitable that ECG data from these devices will be used for both clinical and research purposes. This fact, combined with clinical monitors such as wearable patch recorders capable of recording every heartbeat over 30 days and implantable monitors which can record the patient's ECG for up to 3 years creates an unmanageable amount of ECG data for healthcare providers. As well as that, there are more than 300 million 10 second 12-lead ECGs recorded each year that require interpretation as part of standard care.

The role of AI

ECGs are the most accurate way to diagnose heart arrhythmias and structural abnormalities, but it is impossible for doctors and nurses alone to go accurately and timely through all this data. To address these challenges Artificial Intelligence (AI) will play a crucial role in the future. AI will help physicians make sense of this massive amount of information, provide better patient outcomes and reduce the resource burden that comes from manual data review. It is important ECG interpretation is correct, ensuring timely and sufficient treatment of the patient. AI will allow for effective triaging of ECGs as they come in from the devices, providing a layer of automation between the devices and the physicians, reducing the amount of data that needs to be analysed manually each day.

At PulseAI, we have developed AI which is capable of interpreting ECG data from both clinical devices and consumer devices. Our technology has been developed on a proprietary database of more than 1,000,000 ECG recordings and can accurately detect more than 100 different heart arrhythmias and structural abnormalities, providing a valuable resource to healthcare providers when triaging large amounts of data from their patients. In the future, we hope to see our technology being used in hospitals and clinics across the world, helping to reduce the ECG data burden allowing for more accurate and scalable ECG interpretation.