A comparison study of size-specific dose estimate calculation methods

作者：Parikh Roshni A.; Wien Michael A.^*; Novak Ronald D.; Jordan David W.; Klahr Paul; Soriano Stephanie; Ciancibello Leslie; Berlin Sheila C.

来源：Pediatric Radiology, 2018, 48(1): 56-65.

DOI：10.1007/s00247-017-3986-7

• Children
• Computed tomography
• Radiation dose
• Size-specific dose estimate

Background The size-specific dose estimate (SSDE) has emerged as an improved metric for use by medical physicists and radiologists for estimating individual patient dose. Several methods of calculating SSDE have been described, ranging from patient thickness or attenuation-based (automated and manual) measurements to weight-based techniques.
Objective To compare the accuracy of thickness vs. weight measurement of body size to allow for the calculation of the size-specific dose estimate (SSDE) in pediatric body CT.
Materials and Methods We retrospectively identified 109 pediatric body CT examinations for SSDE calculation. We examined two automated methods measuring a series of level-specific diameters of the patient's body: method A used the effective diameter and method B used the water-equivalent diameter. Two manual methods measured patient diameter at two predetermined levels: the superior endplate of L2, where body width is typically most thin, and the superior femoral head or iliac crest (for scans that did not include the pelvis), where body width is typically most thick; method C averaged lateral measurements at these two levels from the CT projection scan, and method D averaged lateral and anteroposterior measurements at the same two levels from the axial CT images. Finally, we used body weight to characterize patient size, method E, and compared this with the various other measurement methods. Methods were compared across the entire population as well as by subgroup based on body width.
Results Concordance correlation (rho(c)) between each of the SSDE calculation methods (methods A-E) was greater than 0.92 across the entire population, although the range was wider when analyzed by subgroup (0.42-0.99). When we compared each SSDE measurement method with CTDIvol, there was poor correlation, rho(c)< 0.77, with percentage differences between 20.8% and 51.0%.
Automated computer algorithms are accurate and efficient in the calculation of SSDE. Manual methods based on patient thickness provide acceptable dose estimates for pediatric patients < 30 cm in body width. Body weight provides a quick and practical method to identify conversion factors that can be used to estimate SSDE with reasonable accuracy in pediatric patients with body width >= 20 cm.

• 出版日期2018-1