ABSTRACT
Aortic stenosis is a generic condition in aging populations that requires early detection of ventricular decompensation for effective treatment. This condition affects a significant portion of the population, particularly those aged 75 years and older. If not treated properly, severe aortic stenosis can lead to frequent heart failure events and even life-threatening complications such as pulmonary edema and cardiogenic shock that are unresponsive to medical treatment. The current standard treatment for aortic stenosis is aortic valve replacement, which is recommended for symptomatic patients or those with reduced left ventricular ejection fraction. However, it can be challenging and difficult to determine whether aortic valve intervention will improve symptoms and quality of life in patients with comorbidities or multifactorial dyspnoea. Recent advancements in medical treatment for aortic stenosis have provided alternative options for high-risk patients who may not be suitable candidates for aortic valve replacement. Also, it has its limitations, including the need for advanced scanning and interpretation expertise, which may not be readily available in all healthcare settings. Traditional diagnostic methods for aortic stenosis, such as transthoracic echocardiography, have limitations that can result in inaccurate or delayed diagnosis. These advanced techniques have the ability to analyze complex and nonlinear electrocardiogram data, identifying subtle changes in ECG that may indicate the presence of aortic stenosis. The aim of this study is to aid in the early identification of patients at risk of developing aortic stenosis based on European Society of Cardiology (ESC)/European Association for Cardio-Thoracic Surgery (EACTS)/American College of Cardiology (ACC) guidelines along with artificial intelligence (AI), machine learning (ML), and deep learning (DL) approaches.
