Building a quality infrastructure in a large radiography practice

Purpose: Radiographic imaging usually represents the highest volume modality in Radiology. Due to diverse equipment, acquisition challenges, complex image processing, and disparity in technologist training in digital radiography (DR), image quality may not always be optimal. This study aims to report a quality infrastructure for imaging improvement and equipment management in a large radiography practice.

Methods/Materials: Three quality improvement efforts have been established since 2019, including quantitative clinical image scoring, DICOM analytics, and equipment scoring/ranking. Clinical image scoring employs a four-point (0-3) scale based on five categories (patient positioning, centering, collimation and electronic shuttering, labeling, and exposure factors), generating an ideal gold-image score of 15. A Tableau dashboard was created to monitor the image scores and showcase technologists with top scores monthly. DICOM analytics utilizes a set of custom MATLAB programs to extract and compile acquisition- and processing-related DICOM tags of all DR images from 38 radiographic systems. Last, equipment scoring/ranking has been conducted for all radiographic and fluoroscopic (R/F) equipment in the practice plus regional sites. A scoring system was developed based on equipment age, criticality, patient volume per year, parts and service availability, depreciation, and technology relevance. The ranked list has been updated and reviewed on Tableau semi-annually.

Results: Clinical image scoring was conducted for over 52,000 images, with an average score of 13.3. The number of gold images increased from a baseline of 8.2% to 15.0%. DR DICOM analytics from over 2.6 million images provided data for systematic image acquisition investigation, processing harmonization, exposure index (EI) monitoring and EI target optimization, detector utilization, and repeat/reject analysis. Equipment scoring/ranking was successfully implemented for 190 R/F systems initially. The maintained ranked list is used for equipment management and replacement decision making.

Conclusions: A quality infrastructure was built in a large radiography practice and has proven invaluable for image quality improvement and equipment management.