MONTE CARLO SIMULATIONS AS A TOOL FOR DEVELOPING PRE-CLINICAL INSTRUMENTATION

Abstract

Introduction

Nuclear medicine imaging techniques use unsealed radiation sources to obtain images displaying specific physiologic processes, represented by the concentration of emissions in a given volume. In preclinical research, those techniques are crucial for developing new radiopharmaceuticals and characterizing disease models. However, clinical imaging systems are often unsuitable for small animal organs, as their size is typically comparable to the equipment's resolution. While dedicated preclinical instruments exist, their high-cost limits accessibility. Currently, in Brazil, only six centers operate such equipment, mostly for Positron Emission Tomography imaging, with just one operational Single Photon Emission Computed Tomography (SPECT) system in the state of São Paulo.

Objective

To use Monte Carlo simulations to support the development of upgrading strategies for clinical instruments, enabling SPECT and planar imaging of small animals. Simulations will be used to characterize the system's point spread function (PSF) and assess attainable spatial resolution using a digital micro Derenzo phantom. Furthermore, these simulations allow for testing the reconstruction techniques under development.

Materials

The proposed system geometry was simulated using GATE (Geant4 Application for Emission Tomography). It consists of a lead box with pinholes drilled on one lateral side, which collimate photons onto a clinical system's crystal detector. Pinhole magnification helps to overcome spatial resolution limitations. Point sources in different locations inside the FOV were simulated to characterize the system's response. Additionally, to evaluate tomographic response, sets of 40 projections (2 secs each) of a Derenzo phantom containing 10 MBq of 99mTc or 131I were simulated. Images were reconstructed using in-house software, and spatial resolution was assessed based on the smallest group of resolvable rods.

Results

FOV analysis showed that the system's PSF varies with position, with distortions at the edges where points approximated an ellipse before disappearing. This suggests that future reconstruction algorithms must account for these variations. Tomographic reconstruction demonstrated that the current version of the proposed system and algorithm achieve a spatial resolution of around 1.5 mm, suitable for small animal imaging. Future work will focus on building a physical prototype for clinical gamma cameras and comparing experimental data against these simulations.

Conclusions

We present a cost-effective solution for imaging small animal organs by adapting a clinical gamma camera. Furthermore, we developed reconstruction algorithms capable of generating tomographic images. For 99mTc, we observed spatial resolutions of ∼1.5 mm, confirming the system suitability for pre-clinical imaging. Additionally, we demonstrated the capability to acquire images at higher energies (such as from 131I), enabling potential theranostic applications.