A new model for AI can accurately detect, using a twelve-lead electrocardiogram, the patients’ risk of disease progression, worsening disease, or even early death.
It would help the doctors to diagnose an illness at an earlier stage and administer treatment to the worst cases without fail. The paper is titled “Artificial intelligence-enabled electrocardiogram for mortality and cardiovascular risk estimation: It’s an actionable, explainable and biologically plausible platform.
An electrocardiogram (ECG) is a test used to measure the electrical impulses the heart produces and is the most frequently performed diagnostic test in the world.
To provide this, the team used extremely big data from international repositories in the order of millions of different ECGs that have taken in the context of regular health checks to train their AI model to look at an ECG have an understanding of who of the patients on it subsequently developed new disease or had worse disease or died.
Electrical cardiograms represent the electric activity inside as well as between the heart components – the atriums and ventricles. That information was to be fed to the AI model and for ‘it’ to understand or ‘read’ it and to find out patterns in the signals received electrically.
It is believed that the model is able to contemplate ECG patterns discerning more nuanced characteristics than a cardiologist and of greater sophistication.
The researchers also collected imaging and genetics, and these helped them detect that the AI’s predictions correlated with actual biological developments in the heart tissues.
This is one of the things they say is important in the validation of the model with clinicians, demonstrating its ability to detect emerging changes in the structure of the heart overtime that represent initial signs of risk of disease or death.
The principal investigator of this study is Dr Fu Siong Ng, a Reader in Cardiac Electrophysiology, an academic at the National Heart Lung Institute at Imperial College London, a consultant cardiologist at Imperial College Healthcare NHS Trust and a consultant at Chelsea Westminster Hospital NHS Foundation Trust.
Imperial College Healthcare NHS Trust and Chelsea and Westminster Hospital NHS Foundation Trust have already planned trials of AIRE in the NHS. These clinical trials will be implemented to demonstrate the effectiveness of applying the mentioned model on actual clients and will commence in the middle of year 2025. Participants will be selected from OPD, and also from IP medical wards from both the hospitals.
ECG – test has been applied to evaluate the state of the heart for over a century and this study has provided the example of how using artificial intelligence to analyze a perfectly ordinary non-invasive study can bring vital information to the medical world.
‘It may also extend the ability to use ECGs to what has been done before by helping to evaluate risk of future cardiac and health problems, and risk of mortality.’
According to Dr. Sau, there remains “critical need for testing the model in a live health care system.” However, the future can be that patients get such wearable technology which will allow doctors constant and remote control as well as a potential for an alert system from the patient.
Reference:
Sau A, Libor Pastika, Sieliwonczyk E, Konstantinos Patlatzoglou, Ribeiro AH, McGurk KA, et al. Artificial intelligence-enabled electrocardiogram for mortality and cardiovascular risk estimation: The Lancet Digital Health
