Room: Exhibit Hall | Forum 1
Purpose: To develop a method of identifying examinations with deviating technique parameters in an aggregation of Computed Tomography (CT) Radiation Dose Structured Reports (RDSR). Further, to exclude the deviating examinations from calculations of typical dose indices on an acquisition protocol and series level to avoid skewing the results.
Methods: Initially, all RDSR data is sorted by acquisition protocol. For each acquisition protocol, dose indices (CTDIvol and DLP) from each series are binned by combinations of technique parameters found in the RDSR data (e.g., nominal total collimation width, tube rotation time, tube voltage, etc). The series level bins with the highest frequency of usage are assumed to represent the standard acquisition protocol. The remaining series level bins are considered deviating examinations which can be reviewed in a structured way. This method is applied to a CT Chest Abdomen protocol in a trauma setting that uses automatic tube voltage selection (where 120 kV is used for larger patients). Calculations of the median CTDIvol and DLP with all RDSR data are compared to calculations of typical dose indices where the method developed in this work has been applied.
Results: 345 RDSR objects were aggregated for the CT Chest Abdomen protocol. Using all RDSR data, the median CTDIvol and DLP were 10.6 mGy and 395.8 mGycm, respectively. Excluding deviating examinations, the series level (auto-kV selection) dose indices were less for 100 kV (CTDI: 5.8 mGy, DLP: 414.2 mGycm, n: 97) compared to the 120 kV series (CTDI: 13.8 mGy, DLP: 1006.7 mGycm, n: 176).
Conclusion: Typical dose indices can be skewed by examinations that deviate from typical technique parameters. The method developed in this work can help identify deviating examinations so that they can be excluded from calculations of typical dose indices as well as structure the results on a series level.
Funding Support, Disclosures, and Conflict of Interest: Patrik Nowik is a consultant lecturer for Siemens Healthcare Sweden. The other authors have no conflicts of interest to disclose.