Associated particle neutron elemental imaging in vivo: A feasibility study

The purpose of this study is to develop a Monte Carlo simulation model for in vivo associated particle neutron elemental imaging (APNEI) and to study the feasibility of using APNEI to determine the iron distribution in a human liver with the defined model.

The model presented in this study was defined in mcnp by the basic geometry of the human body, the use of D + D source neutrons, iron as the element of interest, an iron-containing voxel in the liver as the target region, and 2 large germanium detectors anterior and posterior to the trunk of the body. The f8 pulse height tally was employed in mcnp to determine the signal acquired from iron inelastic scatter gamma rays at various iron concentrations in the target liver voxel. Correspondingly, the f4 average flux tally in mcnp was modified by a dose function such that the equivalent dose to the whole liver and the effective dose to the whole body could be estimated and used as the basis for a limiting number of neutron histories which could feasibly allow for the collection of a sufficient volume of data to construct a 2D image of iron distribution in the liver voxel.

Assuming an allowable equivalent dose to the liver of 5 mSv, 143 inelastic scatter iron gamma ray counts (at ∼847 keV) would ideally be registered at the germanium detectors for a 1 cm³ cube-shaped liver voxel with an iron concentration of 1000 ppm. According to the simulation model, an image of iron distribution in the liver can be constructed with a 1 cm resolution at the level of 1000 ppm iron. Collecting such an image would yield an estimated whole body dose of 0.82 mSv. The mathematical introduction of image uncertainty resulting from source spot diameter and detector timing resolution more closely approximates the result of real world application.

APNEI of certain elements in vivo appears feasible given several timing, sensitivity, and resolution caveats. However, further study is required to determine what the detection limit of iron would be and what image resolution would be in an experimental setup as the present model contains idealized assumptions which overestimate the signal attributable to iron inelastic scatter gamma rays.