Purpose: Significant anomalies in scanner reported values of CTDIvol for stationary table protocols (with aberrant, elevated values of CTDIvol up to 300 % higher than the actual dose to the phantom), have been observed; well beyond the typical accuracy expected of CTDIvol as a phantom dose. Recognition of these outliers as bad data is important to users of CT dose index tracking systems (e.g. ACR DIR), and a method for recognition and correction is provided. A second anomaly is that the reported CTDIvol for automatic tube current modulation (TCM) has a different physical meaning from the traditional CTDIvol at constant mA, resulting in the dichotomy CTDIvol of the first and second kinds. Methods: Rigorous equations are derived to describe the accumulated dose distributions for auto TCM and for stationary table CT. A comparison with formulae for scanner reported values of CTDIvol clearly identifies the source of these anomalies. Results: For the stationary table, use of the CTDI100 formula (applicable to a moving phantom only) generates extra scatter and an over beaming correction, both of which are non existent if the phantom is held stationary. It is shown that CTDIvol is not robust between fixed mA and auto TCM in this case only DLP is robust. Conclusion: For the case of a stationary phantom, correction of CTDI100 prior to its use in the calculation of CTDIvol using the approach to equilibrium formula H(a) serves as a method of recognition of bad data and as a correction factor for same. In the case of auto TCM, the use of CTDIvol is compromised, and an alternative (approximate) physical interpretation is suggested. However, the reported DLP value (integral dose) remains robust between the fixed and auto mA modes.
ASJC Scopus subject areas
- Radiology Nuclear Medicine and imaging