Authors: Prof. Abhishek Dubey, Akshada Kale, Kashish Mahobiya, Kirti Thakur, Nikita Raj
Abstract: Clinical Decision Support Systems (CDSS) play a crucial role in helping healthcare professionals make accurate, timely, and evidence-driven decisions. However, the growing scale, speed, and diversity of healthcare data have revealed the limitations of traditional rule-based CDSS, especially when dealing with multimorbidity and personalized treatment. Recent advancements in artificial intelligence (AI)—including machine learning, deep learning, and natural language processing (NLP)—have enabled the development of intelligent CDSS that support adaptive learning, predictive analytics, and patient stratification. This paper provides a comprehensive, system-level review of AI-powered CDSS, examining their historical development, underlying technologies, architectural frameworks, and clinical applications. Unlike earlier surveys that focused mainly on individual algorithms, this review integrates AI methods with system architecture, clinical workflows, and ethical considerations. It explores key AI techniques for patient stratification, deep learning models for diagnosis and prognosis, and NLP-driven early warning systems. The paper also addresses critical challenges related to ethics, legal concerns, and explainability, while highlighting emerging trends such as federated learning, digital twins, and genomic-based CDSS. Overall, it aims to offer researchers and clinicians a thorough understanding of AI-CDSS design principles and their future potential.
DOI: http://doi.org/10.5281/zenodo.20706812
