Artificial Intelligence-Based Healthcare Technology: A Semantic Literature Review on Disease Diagnosis and Prediction Systems for Heart Disease, Diabetes, and Cancer Using Machine Learning and Deep Learning Algorithms

Abstract

The rapid advancement of Artificial Intelligence (AI) has significantly influenced the development of intelligent healthcare systems, particularly in disease diagnosis and prediction. Machine learning and deep learning techniques have been widely applied to analyze complex medical data, enabling improved diagnostic accuracy and early disease detection. Despite extensive research in this area, existing studies are often fragmented, focusing on specific diseases or algorithms, which limits comprehensive understanding and cross-domain comparison. This study presents a semantic literature review of AI-based healthcare technologies for disease diagnosis and prediction, with a focus on heart disease, diabetes, and cancer. The review systematically analyzes recent peer-reviewed studies published within the last two years, examining employed datasets, machine learning and deep learning algorithms, evaluation metrics, and application contexts. A semantic categorization framework is adopted to identify relationships among disease domains, data types, algorithmic approaches, and performance indicators. The results reveal prevailing research trends, commonly used models, and emerging methodological practices, including the integration of hybrid models, visualization-based evaluation, and explainable AI techniques. Furthermore, this study highlights existing research gaps and challenges related to data heterogeneity, evaluation standardization, and real-world clinical applicability. The findings provide a structured overview of current advancements and offer valuable insights for future research and development of robust AI-driven healthcare systems.