An Innovative algorithm framework for cardiovascular risk assessment based on ECG data

Background:Cardiovascular disease (CVD) is a primary universal physical problem, with conventional prediction systems frequently being persistent and expensive. Modern advancements in machine learning (ML)offer a hopeful option for accurate CVD risk assessment by leveraging multifaceted relations among diverse risk factors. Aim:Their search proposes a novel deep learning (DL) system, Dynamic Owl Search algorithm-driven Adaptive Long Short-Term Memory (DOS-ALSTM), to enhance cardiovascular risk prediction utilizing electrocardiogram (ECG) data. Method:The study utilizes ECG data from a diverse population group to train and assess the proposed model. Data is cleaned and normalized employing standard techniques to handle lost values and ensure reliability. Relevant features are extracted using statistical and signal processing technique to detain crucial features from the ECG data. The DOS-ALSTM system integrates a DOS optimization algorithm for optimized parameter regulation and ALSTM networks to detain sequential dependencies in ECG data for accurate risk prediction. The recognized method is evaluated using Python software. Result:The DOS-ALSTM system demonstrates superior performance with superioraccuracy of 99%, recall of 98%, F1-Score of 97.9% and Precision of 98.8% in CVD risk assessment compared to traditional methods