Comparing CT Scan Organ and Effective Dose Estimates from Radiation Dose Index Monitoring Systems (RDIMs) across Institutions Using a Reference Set of Patient Datasets
Abstract
Purpose
Radiation Dose Index Monitoring Systems (RDIMs) can report and track estimated organ and effective doses from CT scans. However, the methods to estimate these doses are not standardized. The purpose of this initial study was to assess the feasibility of performing direct comparisons of organ and effective dose estimates from several RDIMs using a reference set of patient datasets.
Methods
Under IRB approval, seven deidentified reference datasets were created and shared across sites: 4 routine thoracic scans and 3 routine abdomen/pelvis CT scans. For each exam, the dataset included: (a) Planning CT Radiograph; (b) Cross sectional CT image data; (c) Radiation Dose Structured Report (RDSR) and (d) Patient Protocol Page. Four sites used their RDIM to produce organ and effective dose estimates for each case. Estimates were compared by calculating the mean value of each organ and effective dose across RDIMs and then the average absolute percent difference (AAPD) across estimates. This was repeated for each estimated organ and effective dose for each patient.
Results
Estimates were provided from two different RDIMs (three sites had the same RDIM product). Mean effective dose estimates across datasets ranged from 6.88 to 12.14 mSV and AAPD of effective doses ranged from 0.1 to 22.9%. The AAPD for key organ dose estimates were: (a) 9.5 to 25.7% for breast and lung doses in thoracic scans; and; (b) 2.5 to 13.9% for colon and stomach doses in abdomen/pelvis scans. For the three sites using the same RDIM product, AAPDs were < 3% for key organ dose estimates and < 4% for effective dose estimates, respectively.
Conclusion
This study demonstrated the feasibility of performing direct comparisons of organ and effective dose estimates using a reference set of patient datasets. Future work will investigate sources of observed differences in RDIM results and possible mitigation strategies.