Date/Time
Date(s) - 10/11/2012
12:30 pm - 1:30 pm

Abstract:
Computational phantoms are computerized representations of human anatomy for use in radiation transport simulation of either internal or external radiation exposures.   For medical applications, they are needed for modeling organ doses resulting from medical image acquisition or from radiation therapy procedures.  In this paper, we will review the development, status and application of a series of hybrid phantoms created at the University of Florida.  Methods:  CT and MR images of adult, pediatric and fetal subjects were acquired and used to create an array of ICRP 89 compliant hybrid phantoms based upon combinations of polygon-mesh and NURBS surfaces.  For the post-natal models, phantom-specific skeletal models were assembled based upon either 3D microCT images of trabecular spongiosa or previously acquired linear pathlength distributions in pediatric bone.  Values of intra-skeletal electron absorbed fractions were then assembled and used to generate photon fluence-to-dose response functions for reporting photon dose to active and shallow marrow during radiation transport.  Data on body morphometry characteristics of the current U.S. population were further reviewed and used to design an expanded library of patient-dependent adult and pediatric hybrid phantoms for patient-phantom matching in medical dosimetry.  Results:  Reference and patient-dependent phantoms of the UF series have been applied to clinically relevant studies of organ dosimetry for (1) nuclear medicine, (2) computed tomography, (3) interventional fluoroscopy, and (4) radiation therapy in both retrospective and prospective patient studies.  Conclusions:  Since the mid-2000s, significant advancements in phantom-based dosimetry studies have been made by many investigative teams using hybrid phantom technology.  In situations where medical images of the individual patient are either unavailable or cover only a portion of the relevant anatomy, phantom-based assessments of organ dose are essential for characterizing risks associated with therapy or diagnostic imaging procedures.  Advances in patient-specificity via patient-dependent hybrid phantom libraries can significantly lower dose uncertainties over traditional methods based upon fixed-anatomy stylized or even voxel-based models.

Biosketch:
Dr. Bolch is currently a Professor of Biomedical Engineering at the University of Florida and Director of ALRADS – the Advanced Laboratory for Radiation Dosimetry Studies at UF.  Dr. Bolch manages a broad research program including (1) NCI and DOE funded projects to construct high-resolution models of the skeleton to support dose-response studies in radionuclide therapy and radiation epidemiology, (2) NIBIB funded projects to develop scalable NURBS-based and voxel-based computational phantoms of adult and pediatric patients for organ dose assessment in CT, fluoroscopic imaging, and radiotherapy, and (3) CDC funded projects in stochastic modeling of worker inhalation and gamma-ray exposures from radiological accidents and potential terrorist events.  He is member of the Society of Nuclear Medicine’s Medical Internal Radiation Dose (MIRD) Committee since 1993, Member of Committee 2 of the International Commission on Radiological Protection (ICRP) since 2005 (Secretary as of 2012), and member of the National Council on Radiation Protection and Measurements (NCRP) since 2005.  In 2012, he was elected Fellow of both the American Association of Physicists in Medicine and of the Health Physics Society.