Medical physicist J Anthony Seibert and his colleagues from the Department of Radiology described a situation in which a 14-year-old patient had two CT scans, and a four-fold difference in CTDIvol was recorded between the two. The exams – a scan to help plan surgery and a post-surgical scan ordered for clinical suspicion of pneumonia – were performed using two different CT modalities (Definition 128 AS+, Siemens and VCT, GE Healthcare). The CTDIvol and DLP values were given to the child's parent without any calculations made to estimate radiation dose. One set of values was four times higher than the other. The alarmed parent wanted to know the reason for the "overdose".

The medical physicists who investigated this case determined that the child actually received approximately the same dose in both scans. They reached this conclusion by calculating SSDE values, using the methodology published in the 2012 AAPM Task Group 204 "Size Specific Dose Estimates in Pediatric and Adult CT Examinations". This publication provides conversion factors as a function of effective patient diameter for the 32 and 16 cm diameter CTDIvol calibration phantoms.

"Because of differences that can exist between patient effective diameter and calibration phantom diameter, the CTDIvol referenced to the patient can be significantly under- or over-estimated. Dose estimates that rely on CTDIvol values reported by a CT scanner can vary widely, even when the actual patient dose is comparable," the authors explained.

Although not perfect, the use of SSDE methods can significantly reduce discrepancies between the patient and calibration phantom size, and can also account for the CT manufacturers' choice of CTDIvol calibration phantom diameter.

SSDE is the first initial step in a long-range process by the AAPM CT subcommittee to realize more accurate patient-specific dose and risk estimates, Seibert told medicalphysicsweb. It is currently defined only for the torso, and is not applicable for head CT scans because the research has not yet been conducted.

Seibert also pointed out that SSDE doesn't modify the DLP (defined as CTDIvol multiplied by table travel for the scan). "Many might assume that the SSDE multiplied by the table travel is a better estimate for the DLP, but research and evaluation have not yet been conducted, so the DLP remains unchanged at this point."

"Because the SSDE takes into account the patient's effective diameter, the radiation dose to the patient can be more accurately estimated based upon the reported CTDIvol and the corresponding adjustment factors," he said. "It was serendipitous that my colleagues were working on the TG-204 document when we received the request to review the case of the patient."

"The ACR Dose Index Registry now estimates patient effective diameter from the CT localizer, and according to TG-204 look-up tables, calculates SSDE for each exam. In addition, CT manufacturers and radiation dose software vendors are now implementing methods to determine 'water-equivalent' effective patient diameter directly from axial scans, and coupled with CTDI phantom diameter and CTDIvol information, SSDE is calculated. However, common use of SSDE is still some time away."

For now, Seibert and colleagues recommend using SSDE when requested to calculate dose estimates of CT chest and abdomen exams, especially when the patient's body size does not conform well to either of the standard CTDI phantom sizes.

Related stories

• Dose coefficients simplify CT calculations
• CT Gently app offers scan optimization
• CT lung dose cut to 0.25 mSv
• Phantoms assess CT dose to obese patients