Designing a Real-Time Human-AI Decision Support Framework for U.S. Healthcare Providers Using Big Data Analytics

Abstract

Clinical decision-making urgency and the growing complexity of healthcare data sets have necessitated a critical and dire necessity of intelligent systems capable of providing support to healthcare providers in the U.S. in real-time. In this work we suggest developing a Human-AI decision support system based on big data analytics to improve the quality of diagnosis, cognitive overload, responsiveness rate in the context of high-stakes medical settings. Using the methodologies behind the AI-enhanced software quality assurance, predictive analytics, and digital twin integration, as illustrated by Joy, Alam, Bakhsh, and their colleagues, the present research makes business-critical software testing frameworks applicable to the healthcare sector. The architecture provided suggests a real-time ingestion of data, performance of a predictive model, AI-based decision support, as well as a human-in-the-loop check of the results to guarantee correctness and responsibility. It was based on a qualitative approach, comprising comparative framework analysis, healthcare-related use case mapping, and cross-domain provision of AI-collaboration models initially designed within the scope of QA-BA ecosystems. In terms of key findings, it has been indicated that such a hybrid model can be very useful when it comes to the enhancement of the reliability of decisions made, especially in the case of acute care alongside with opportunities towards continuous learning and agile workflow. In this work, the positive transformative power of AI-human synergy in the context of healthcare is singled out and the basis is prepared to implement intelligent clinical support systems that meet ethical, operating, and technical limitations of the U.S. healthcare organizations in the future.