Heart It has a chronological age (that of the person) but, according to its functioning, it also has a ‘biological’ age. Thus, someone of 50 years with poor heart health can have a biological age of 60 years, and another of 50 with good heart health can have a biological of 40.
Knowing it is important to prevent diseases and to identify people with the highest risk of cardiovascular events and mortality.
Today, a team of researchers from South Korea has presented at the Scientific Congress of the European Society of Cardiology (ESC) An algorithm that predicts the biological age of the heart.
The equipment has achieved it after using artificial intelligence (AI) to analyze standard electrocardiograms (ECG) data of 12 derivations (allows to visualize the electrical activity of the heart) of half a million people.
“The investigation showed that when the biological age of the heart exceeded the chronological age seven years, the risk of mortality due to all causes and major adverse cardiovascular events increased sharply”, Avanza Yong-Soo Baek, of the Inha University Hospital in South Korea.
“On the contrary, if the algorithm estimated that the biological heart was seven years younger than chronological age, the risk of death and major adverse cardiovascular events was reduced.”
Besides, “Use AI to develop algorithms in this way introduces a possible paradigm shift in cardiovascular risk assessment”, Points out the researcher and main author of the study.
Predictive power against chronological age
The study evaluated the prognostic capacities of an algorithm based on deep learning that calculates the cardiac age of the biological ECG (cardiac age of the ECG AI) from ECG of 12 derivations, comparing its predictive power against the traditional chronological age for mortality and cardiovascular results.
A deep neuronal network was developed and trained in a substantial data set of 425,051 ECG of 12 derivations collected for fifteen years, and then validated in an independent cohort of 97,058 ECG and comparative analysis between patients of the same age and sex were performed.
In statistical models, An age of the heart in the ECG of the upper AI in seven years at the chronic age of the heart was associated with an increased risk of mortality by all causes of 62% And, on the contrary, an age of the heart in the ECG AI seven years lower than its chronological age reduced the risk of mortality by all causes by 14%.
However, Baek warns, “it is crucial to obtain a statistically sufficient sample size in future studies to corroborate these findings. This approach will improve the solidity and applicability of the ECG of AI in the clinical evaluations of heart function and health.”
“The estimated biological cardiac age by artificial intelligence from electrocardiograms of 12 derivations is strongly associated with an increase in mortality and cardiovascular events, which underlines its usefulness to improve early detection and preventive strategies in cardiovascular health care,” says Baek.
This study confirms the transforming potential of the AI to improve clinical evaluations and improve the results of the patients, the authors conclude.