A Review of Semantic Literature on Stroke, Mental Health, and Predictive Genomics: Artificial Intelligence in Medical Diagnosis

Abstract

Artificial Intelligence (AI) has increasingly become a transformative force in modern healthcare by enabling early disease detection, improving diagnostic accuracy, supporting personalized treatment strategies, and enhancing data-driven clinical decision-making. This study presents a semantic literature review that systematically examines and synthesizes existing research on the application of AI in medical diagnosis across three high-impact and clinically significant domains: stroke and ischemic brain events, mental health disorders, and predictive genomics, including neurodegenerative diseases such as Parkinson’s disease and complex genetic conditions like age-related macular degeneration (AMD). Rather than proposing a novel algorithmic model, this review focuses on analyzing and categorizing prior studies that employ machine learning (ML), deep learning (DL), and hybrid AI approaches for disease diagnosis and prediction. A total of 25 internationally peer-reviewed journal articles published between 2015 and 2025 were selected based on predefined inclusion criteria, emphasizing methodological rigor, dataset characteristics, algorithmic performance, and clinical relevance. The semantic review approach allows for the identification of thematic patterns, comparative trends, and knowledge gaps across the selected medical domains. The findings indicate that AI-based diagnostic systems demonstrate strong potential in improving diagnostic sensitivity and specificity, particularly in stroke imaging analysis, mental health disorder classification using behavioral and neuroimaging data, and genomics-based risk prediction. However, challenges related to data heterogeneity, model interpretability, dataset bias, and clinical integration remain significant barriers to widespread implementation. This review highlights emerging research directions and future opportunities for AI-assisted diagnostics, emphasizing the need for explainable AI, standardized datasets, and interdisciplinary collaboration between clinicians and data scientists. Overall, this study provides a comprehensive overview of current advancements and limitations of AI in medical diagnosis, offering valuable insights for researchers, healthcare practitioners, and policymakers.