Scholarlink

This study investigates the advancements in heart disease identification through the application of artificial intelligence (i.e., machine learning (ML) and deep learning (DL) methods. The main goal is to enhance early detection and prediction accuracy by utilizing a range of ML and DL models and analyzing heterogeneous datasets. Key methods, including feature selection, data preparation, and model optimization, are highlighted as important to improve the cumulative performance and dependability of the models. Heart disease, which affects the coronary arteries that provide blood to the heart, remains the leading cause of global mortality, with approximately 18.2 million fatalities yearly, based on the Centers for Disease Control and Prevention (CDC). ML and DL techniques have the potential to assist at the outset of detection and diagnosis of heart disease (HD) before it progresses to a critical and challenging stage. This study employs these techniques to diagnose HD by analyzing various physical and health attributes such as age, cholesterol level, chest pain, and more. The relationship between these attributes and their contribution to HD prediction is explored through comprehensive data analysis. Rapid detection of HD is importantly vital in lowering death rates and improving individuals’ quality of life, thus obviating serious harm such as cardiac arrest, stroke, and disability. To evaluate the effectiveness of various algorithms in diagnosing HD, evaluation metrics such as accuracy, confusion matrix, and ROC curve are utilized. Among the models tested, logistic regression (LR) achieves the highest accuracy of 84.02% using ten-fold cross-validation. The DNN stands out particularly in terms of precision, sensitivity, and F-measure, 90.03, 87.14, and 88.08, respectively. The findings demonstrate that combining ML and DL techniques significantly enhances diagnostic capabilities, providing healthcare practitioners with valuable tools for more accurate diagnosis and treatment of cardiac disease. This work contributes to the advancement of diagnostic technologies, ultimately improving patient outcomes.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Improving Heart Disease Detection with Data Analysis: From Machine and Deep Learning Perspective

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Lateef Adesola Akinyemi,
  • Jelil Olatunbosun Agbo-Ajala,
  • Olufisayo Sunday Ekundayo,
  • Donatien Koulla Moulla,
  • David Attipoe,
  • Sree Ganesh Thotempudi,
  • Ernest Mnkandla,
  • Mbuyu Sumbwanyambe

摘要

This study investigates the advancements in heart disease identification through the application of artificial intelligence (i.e., machine learning (ML) and deep learning (DL) methods. The main goal is to enhance early detection and prediction accuracy by utilizing a range of ML and DL models and analyzing heterogeneous datasets. Key methods, including feature selection, data preparation, and model optimization, are highlighted as important to improve the cumulative performance and dependability of the models. Heart disease, which affects the coronary arteries that provide blood to the heart, remains the leading cause of global mortality, with approximately 18.2 million fatalities yearly, based on the Centers for Disease Control and Prevention (CDC). ML and DL techniques have the potential to assist at the outset of detection and diagnosis of heart disease (HD) before it progresses to a critical and challenging stage. This study employs these techniques to diagnose HD by analyzing various physical and health attributes such as age, cholesterol level, chest pain, and more. The relationship between these attributes and their contribution to HD prediction is explored through comprehensive data analysis. Rapid detection of HD is importantly vital in lowering death rates and improving individuals’ quality of life, thus obviating serious harm such as cardiac arrest, stroke, and disability. To evaluate the effectiveness of various algorithms in diagnosing HD, evaluation metrics such as accuracy, confusion matrix, and ROC curve are utilized. Among the models tested, logistic regression (LR) achieves the highest accuracy of 84.02% using ten-fold cross-validation. The DNN stands out particularly in terms of precision, sensitivity, and F-measure, 90.03, 87.14, and 88.08, respectively. The findings demonstrate that combining ML and DL techniques significantly enhances diagnostic capabilities, providing healthcare practitioners with valuable tools for more accurate diagnosis and treatment of cardiac disease. This work contributes to the advancement of diagnostic technologies, ultimately improving patient outcomes.