Fifty-sixth annual meeting of the American association of physicists in medicine
SU-C-12A-02: CTDIvol Variability in Tube-Current Modulated Body CT: The Role of Patient Diameter
Researchers have used the large variability in CTDIvol as an example of inconsistent and inappropriate protocols for CT imaging. However, CTDIvol variability can exist for many sound reasons pertaining to protocol needs (slice thickness, etc.) or patient size. This investigation focused on the role that patient effective diameter plays in CTDIvol variability.
A database associated with a commercial CT dose reporting tool was mined, and the protocol parameters from 3136 patients who were scanned on one CT scanner (GE VCT) at our institution were evaluated. All patients were imaged using an abdomen-pelvis protocol, and tube current modulation was employed. A graph of CTDIvol versus diameter was fit using an exponential function (CTDIvol = a EXP(b x DIAMETER), and the parameters of this fit were used to subtract away the diameter dependence. Evaluation of variance before and after this correction was performed.
Abdomen-pelvis CT scans performed between February 28, 2013 and February 28, 2014 were evaluated corresponding to a timeframe where x-ray tube current modulation was used consistently. Average patient age was 50.0 years (SD=18.1), and ranged from 0 to 97 years. Mean patient diameter was 306.2 mm (SD=52.0), and CTDIvol values ranged from 2.8 mGy to 56.0 mGy, with a mean of 11.1 mGy (SD=6.68 mGy). After the dependency of patient diameter was removed from the data, the remainder data averaged 0.28 mGy (SD=3.01 mGy). A 2.2 fold reduction in standard deviation was observed, and the reduction in variance achieved by removing the size dependence of CTDIvol was (6.68/3.01)∧2 = 4.93.
In a modern CT scanner using contemporary TCM protocols, a large component of CTDIvol variance can be attributed to patient size. The reduction of variance by a factor of 4.9 demonstrates that significant variability in CTDIvol is related to patient size. This study was partially funded by a grant from the University of California Office of the President