The Mallinckrodt Institute of Radiology (MIR) at Washington University School of Medicine is a world-renowned academic department dedicated to clinical excellence, research innovation, and education in medical imaging. As part of a major academic medical center, MIR provides a full spectrum of diagnostic and interventional radiology services while driving forward the scientific frontiers of imaging technologies, including MRI, CT, PET, and molecular imaging. Its mission integrates patient care with groundbreaking research and the training of future radiologists and scientists.

Why AI matters at this scale

For an institute of MIR's size (501-1000 employees), operating at the intersection of high-volume clinical service and cutting-edge research, AI is not a distant future but a present-day imperative. At this scale, manual processes and subjective image interpretation create bottlenecks that affect patient wait times, research throughput, and operational costs. AI offers the leverage to amplify the expertise of its specialists, automating repetitive tasks to free them for complex cases and innovative research. For a mid-sized entity within a larger university system, strategic AI adoption can create disproportionate competitive advantages in research grants, clinical trial recruitment, and reputation, while directly addressing pressures to improve efficiency and patient outcomes.

Concrete AI Opportunities with ROI Framing

1. Diagnostic Workflow Acceleration: Implementing AI-based triage for neurological and thoracic CT scans can reduce the time to identify critical findings like strokes or pulmonary embolisms. The ROI is measured in improved patient outcomes (reduced morbidity) and increased radiologist productivity, allowing more studies to be read per day without adding staff. 2. Enhanced Clinical Trial Imaging Analysis: MIR participates in numerous research studies requiring precise tumor measurements. AI tools for automated segmentation and tracking of lesion changes over time can reduce analysis time from hours to minutes per case. This directly translates to faster trial results, lower labor costs for image analysis, and a more attractive service for pharmaceutical sponsors, generating new revenue streams. 3. Operational Efficiency in Imaging Scheduling & Utilization: An AI model predicting no-show rates for MRI appointments and optimizing scan sequence protocols based on clinical indication can significantly improve equipment utilization. Filling cancelled slots and reducing average scan times increases patient throughput, boosting revenue from existing capital-intensive assets without new hardware purchases.

Deployment Risks Specific to this Size Band

MIR's size presents unique deployment challenges. While large enough to have dedicated IT and research informatics staff, it may lack the vast, centralized data science teams of mega-hospitals, requiring careful prioritization and potential partnership with university resources or external vendors. Budget approvals for enterprise AI platforms may compete with other capital needs like new imaging equipment. Furthermore, integrating AI tools into a heterogeneous technology environment—potentially involving multiple Picture Archiving and Communication Systems (PACS) and Electronic Health Records (EHR)—requires significant IT project management to ensure seamless clinician adoption. Finally, at this scale, proving the value of a pilot project is crucial to secure funding for broader rollout, necessitating a strong focus on measuring and communicating clear metrics related to clinical, operational, or financial performance.