AI for Optimizing Imaging Department

Author(s): Wertheim Ofir, Ben-Zakai Itiel, Rabl Yaniv, Massalja Jammal, Jerdev Michael, Mizrahi Hagar, Erez Onn, Shimon Sabah, Yehudai Noam, Blum Arnon

Objectives: Demand for medical imaging continues to rise while radiology departments face personnel shortages, administrative bottlenecks, and increasing diagnostic complexity. Artificial intelligence (AI) offers tools that can automate repetitive tasks, streamline workflows, and enhance operational efficiency. This review examines contemporary literature on AI applications in imaging department management, with a special focus on the real-world implementation of an AI-assisted referral processing and scheduling system in a hospital CT unit.
Methods: A narrative review of peer-reviewed literature on AI in radiology workflow optimization was conducted, focusing on optical character recognition (OCR), natural language processing (NLP), triage systems, protocol selection, automated scheduling, and human-in-the-loop oversight models. Findings are integrated with a real-world case study from Tzafon Medical Center, where an AI workflow incorporating OCR, NLP, and a deterministic rule engine was deployed to support CT operations.
Results: Published studies demonstrate that AI can reduce administrative burden, improve protocol accuracy, assist triage, shorten turnaround times, and increase imaging throughput. The case example demonstrates comparable gains: automated referral processing improved CT volume by 20%, reduced administrative workload by 10 hours weekly, shortened waiting times by 30%, improved patient satisfaction by ~12%, and reduced annual complaint rates by 95%.
Conclusions: AI-enhanced workflow systems can substantially improve the efficiency, accuracy, and safety of imaging services—particularly when paired with robust human oversight. While clinical decision-making remains primarily the responsibility of trained radiologists and clinicians, administrative AI provides a scalable and reproducible method for optimizing imaging department operations. Future development will likely focus on more adaptive decision support, patient-specific protocol customization, and integration with hospital-wide resource management platforms.