Optimizing Abdominal CT Protocols via Crowd-Sourced Analysis of 908,000 Exams
Overview
A multinational study analyzing 908,000 routine abdomen CT exams revealed a six-fold variation in radiation doses across 1,033 protocols after adjusting for patient size. Using crowd-sourced data and cluster analysis, the study identified optimized protocols that minimize radiation exposure while maintaining diagnostic image quality.
Background
Abdominal CT is the second most common CT examination in Europe and contributes a relatively high effective radiation dose compared to other diagnostic imaging procedures. Despite known dose reduction technologies, there is substantial variation in radiation dose driven primarily by local protocol decisions rather than patient or machine factors. This variation increases patients' excess cancer risk and highlights the need for standardized, optimized CT protocols. However, developing and maintaining such protocols is resource-intensive and complex due to trade-offs between dose, image quality, scan time, and anatomical coverage.
Data Highlights
Parameter
Value
Number of abdomen CT exams analyzed
908,000
Number of unique protocols
1,033
Range of average radiation dose variation
Six-fold
Number of hospitals/facilities
132
Countries involved
7
Data collection period
2015–2021
Key Findings
There is a six-fold variation in average radiation dose across 1,033 unique abdomen CT protocols after adjusting for patient size.
Variation in radiation dose is primarily driven by local protocol decisions rather than patient size or scanner characteristics.
Many facilities underutilize dose reduction technologies such as automated kV modulation and limiting unindicated scan phases.
Crowd-sourced data from a large international registry enables identification of low-dose optimized protocols.
Cluster analysis groups protocols with similar technical parameters, facilitating identification of best practices.
Maintaining and updating CT protocols is resource-intensive, often requiring full-time effort per indication.
Clinical Implications
Clinicians and radiology departments should prioritize protocol standardization and leverage large-scale data to identify and adopt optimized low-dose abdomen CT protocols. Utilizing dose reduction technologies more consistently and limiting unnecessary scan phases can reduce patient radiation exposure without compromising image quality. Collaborative, data-driven approaches can streamline protocol management and improve patient safety.
Conclusion
This large-scale, crowd-sourced analysis demonstrates significant variation in abdominal CT radiation doses driven by protocol choices and highlights the potential for data-driven optimization to reduce unnecessary radiation exposure while maintaining diagnostic quality.
References
University of California San Francisco International CT Dose Registry Study
