Monte Carlo study of a 3D Compton imaging device with GEANT4

TL;DR

This study uses Monte Carlo simulations with Geant4 to evaluate a novel 3D Compton imaging technique for detecting density variations in objects like the brain, achieving high resolution and contrast at low radiation doses.

Contribution

It provides a detailed simulation analysis of a new 3D Compton imaging method, demonstrating its effectiveness for brain imaging and density variation detection.

Findings

Achieves 2 mm resolution for 30% density increase inclusions.

Detects density variations with about 12% contrast.

Estimates a 1 mSv dose for a complete brain scan.

Abstract

In this paper we investigate, with a detailed Monte-Carlo simulation based on Geant4, the novel approach [Nucl. Instrum. Methods A588 (2008) 457] to 3D imaging with photon scattering. A monochromatic and well collimated gamma beam is used to illuminate the object to be imaged and the photons Compton scattered are detected by means of a surrounding germanium strip detector. The impact position and the energy of the photons are measured with high precision and the scattering position along the beam axis is calculated. We study as an application of this technique the case of brain imaging but the results can be applied as well to situations where a lighter object, with localized variations of density, is embedded in a denser container. We report here the attainable sensitivity in the detection of density variations as a function of the beam energy, the depth inside the object and size and density of the inclusions. Using a 600 keV gamma beam, for an inclusion with a density increase of 30% with respect to the sorrounding tissue and thickness along the beam of 5 mm, we obtain at midbrain position a resolution of about 2 mm and a contrast of 12%. In addition the simulation indicates that for the same gamma beam energy a complete brain scan would result in an effective dose of about 1 mSv.