PT - JOURNAL ARTICLE AU - Steven Leung AU - Clara Chen AU - Roberto Maass-Moreno TI - Using tomographic images for internal exposure dose estimates DP - 2013 May 01 TA - Journal of Nuclear Medicine PG - 2100--2100 VI - 54 IP - supplement 2 4099 - http://jnm.snmjournals.org/content/54/supplement_2/2100.short 4100 - http://jnm.snmjournals.org/content/54/supplement_2/2100.full SO - J Nucl Med2013 May 01; 54 AB - 2100 Objectives Dose estimates of internal radiation exposure (IDE) are based on the bio-distribution (BioD) of radiotracers in a standard (normal) individual. However, in many patients, the "standard" BioD model does not always apply. The objective of this study was to find an image-based IDE method, applicable to individual patients, to evaluate the errors caused by assuming standard BioD. The proposed procedure uses acquired 3D images as indicators of actual BioD. Assumptions: 1) Patient images show areas of any significantly abnormal BioD and 2) static 3D images represent BioD in quasi steady state. Methods Pre- and post-treatment PET volumes (PreImg and PostImg, respectively) were anisotropicaly scaled along their three axes to closely overlay the reference MIRD human phantom (RefImg). Organ overlap between the PET volumes and the RefImg provided the segmentation. The values of PreImg were scaled so its integral was equal that of PostImg over the entire common volume. The ratio of organ averages, PreImg/PostImg, was applied as a correction factor for the standard (adult) normalized cumulated activity to re-compute the IDE using OLINDA. Results Segmentation inaccuracies were minimized for this study by using oncologic FDG-PET images pre- and post- successful treatment. In the clearest case of significant bio-redistribution (mesothelioma), if one assumes “standard” IDE for the PostImg, the effective dose increased by +10.5% and the largest individual organ dose change was +48%. Conclusions While the recalculations of IDE may have been affected by the somewhat crude registration method, they revealed local quantitative variations by which an image-based BioD will produce estimates that are more realistic and significantly different from those obtained with the “standard” BioD approach. This is especially evident in situations where radiation sources have significant redistribution or appear near organs with high radiosensitivity. Research Support NIH Clinical Center 2012 Summer Internship Program