South Korean researchers have developed an artificial intelligence (AI)-based algorithm capable of predicting the biological age of the heart, offering an advancement in cardiovascular risk assessment. Presented at the European Society of Cardiology (ESC) scientific congress, the study underscores the importance of understanding the heart's biological age, which can differ substantially from an individual's chronological age.
The team, led by Yong-Soo Baek from Inha University Hospital in South Korea, developed a deep neural network trained on a dataset of 425,051 12-lead electrocardiograms (ECGs) collected over fifteen years. This AI algorithm estimates the biological age of the heart by analyzing standard ECG data, providing insights into the heart's functioning beyond chronological measurements.
"The biological heart age estimated by artificial intelligence from 12-lead electrocardiograms is strongly associated with increased mortality and cardiovascular events, underscoring its utility to improve early detection and preventive strategies in cardiovascular healthcare," Baek said.
The algorithm was validated on an independent cohort of 97,058 ECGs, with comparative analyses performed between patients of the same age and sex. The study evaluated the prognostic capabilities of this deep learning-based algorithm, comparing its predictive power against traditional chronological age for mortality and cardiovascular outcomes.
Findings revealed that when the biological age of the heart exceeded the chronological age by seven years, there was an increase in health risks. "The research demonstrated that when the biological age of the heart exceeded the chronological age by seven years, the risk of mortality from all causes and of major adverse cardiovascular events increased sharply," Baek stated.
In statistical models, an AI ECG heart age exceeding the chronological age by seven years was associated with an increased risk of all-cause mortality by 62 percent and of major adverse cardiovascular events (MACE) by 92 percent. MACE includes critical conditions such as heart attack, stroke, cardiovascular death, and revascularization procedures like angioplasty and bypass surgery.
Conversely, the study found that if the algorithm estimated that the biological heart was seven years younger than the chronological age, health outcomes improved. "On the contrary, if the algorithm estimated that the biological heart was seven years younger than the chronological age, the risk of death and major adverse cardiovascular events was reduced," Baek noted.
In particular, an AI ECG heart age that was seven years younger than its chronological age reduced the risk of all-cause mortality by 14 percent and MACE by 27 percent. This emphasizes the potential benefits of having a biologically younger heart in preventive healthcare strategies.
The concept of biological heart age is based on the functioning of the organ rather than the number of years a person has lived. This biological age is a better predictor than chronological age for the risk of death and major cardiovascular events. For instance, a 50-year-old individual with poor heart health may have a biological heart age of 60 years, while another 50-year-old with good heart health may have a biological heart age of 40 years.
Baek discussed the impact of integrating AI into cardiovascular assessments. "Using AI to develop algorithms in this way introduces a potential paradigm shift in cardiovascular risk assessment," he said. The integration of artificial intelligence in clinical diagnostics offers new opportunities to improve predictive accuracy in cardiology.
The study also explored the link between the AI-estimated heart age and other cardiac parameters. Subjects with reduced ejection fraction, a measure of how much blood the left ventricle pumps out with each contraction, systematically showed increased AI ECG heart ages. These individuals also exhibited prolonged QRS durations, which is the time taken for the heart's electrical signal to travel through the ventricles, causing contraction. These indicators of electrical remodeling within the heart may indicate underlying heart health problems and their association with ejection fraction.
"It is crucial to obtain a statistically sufficient sample size in future studies to further corroborate these findings. This approach will improve the robustness and applicability of AI ECG in clinical assessments of cardiac function and health," Baek emphasized.
The authors of the study concluded that their findings confirm the potential of AI to refine clinical assessments and improve patient outcomes. Knowing the biological age of the heart is important for preventing diseases and identifying people at higher risk of cardiovascular events and mortality, as it helps in predicting these risks based on heart function.
"Cardiac events are difficult to predict but are a major risk of death," the researchers noted. The development of such AI algorithms could pave the way for improved screening and early intervention, potentially reducing the burden of cardiovascular diseases globally.
The article was written with the assistance of a news analysis system.