اولین همایش بین المللی تحول خدمات سلامت: تلفیق مراقبت، درمان، بازتوانی و امیدبخشی

رفتن به صفحه اصلی

ارائه مجازی مقاله

کد مقاله 1215

عنوان AI-Based Clinical Decision Support Systems (CDSS) in Early Disease Diagnosis: A Systematic Review

نویسندگان Ali Akbar Oboudi - Alireza Zarei

چکیده Abstract: Background: The increasing complexity of healthcare data strains traditional diagnostic workflows. While Artificial Intelligence-based Clinical Decision Support Systems (AI-CDSS) offer enhanced diagnostic accuracy and predictive modeling, their clinical integration remains limited. This systematic review aims to evaluate the current landscape of AI-CDSS for early disease diagnosis, identify critical adoption barriers, and bridge the gap between theoretical models and real-world healthcare applications. Methods: Following PRISMA guidelines, a systematic search was conducted across PubMed, Scopus, Web of Science, and IEEE Xplore for articles published between January 2015 and January 2025. Keywords included “Artificial Intelligence,” “Clinical Decision Support Systems,” “Early Diagnosis,” and “Machine Learning.” Peer-reviewed studies focusing on clinical application and algorithmic performance of AI-CDSS were included. Non-English publications, reviews, preprints, and studies lacking clinical evaluation were excluded. Ultimately, 11 high-quality studies were synthesized. Findings: A significant gap exists between high algorithmic accuracy in controlled research and practical healthcare implementation. Critical bottlenecks include algorithmic interpretability (the “black box” problem), data privacy, interoperability with existing Electronic Health Records (EHR), and the need for rigorous empirical validation. Furthermore, evidence suggests that AI models must be tailored to specific clinical workflows rather than relying on generalized predictive solutions. Conclusions: AI-driven CDSS is foundational for proactive, personalized medicine. To overcome the theory-practice gap, institutions must adopt explainable and tailored AI architectures. Resolving interoperability limits, establishing robust ethical frameworks, and ensuring comprehensive clinical validation are essential for successful real-world deployment.

فایل ها

فایل پوستر مقاله

محتوای ارائه مجازی مقاله

نظرات