ABSTRACT
The motivation for producing a pediatric computational phantom was to determine the absorbed organ doses from computed tomography (CT) examinations using the public domain EGS4 Monte Carlo code.1 The project was conceived in about 1992 in response to the dearth of available absorbed dose data for children who underwent CT procedures. Inspiration was derived from the work of Cristy, who extrapolated the mathematical medical internal radiation dose (MIRD) computational phantoms to children2 and from the work of Zankl et al. who produced voxel computational phantoms of two children.3 The Monte Carlo dose calculations for CT procedures using “adult” MIRD computational phantoms by Shrimpton et al.4 and later for plain radiography in children with MIRD computational phantoms5 also inspired the ADELAIDE work. In the early 1990s, desktop PCs were not capable of performing Monte Carlo calculations; picture archiving and communication systems (PACS) were not yet available; the Digital Imaging and Communications in Medicine (DICOM) standard for distributing and viewing medical image from any origin was not operating; computers were not networked; the electronic
transfer of data was limited by the capacity of the “ oppy disk”; and the Internet had not yet developed into the information distributing system that it currently is. Consequently, working with medical images on computers was not straightforward.6 The situation in 2007 is much better and the number of voxel computational phantoms has increased dramatically.7 Despite the improvements in computer hardware and software, segmenting the large number of medical images required for a voxel computational phantom is a timeconsuming task. Even though steps have been taken to semiautomate the task,8-10 segmentation is still a lengthy operation. This chapter describes the segmentation procedure used for ADELAIDE’s construction and the anatomy of the computational phantom.
