Development and Evaluation of a High Spatial Resolution Preclinical SPECT Device based on Large Area Scintillator Crystals, Pinhole Collimators and Artificial Intelligence

• Grant number: 24/09976-5
• Support Opportunities: Regular Research Grants
• Start date: April 01, 2025
• End date: March 31, 2027
• Field of knowledge: Interdisciplinary Subjects
• Principal Investigator: Jorge Mejía Cabeza
• Grantee: Jorge Mejía Cabeza
• Host Institution: Instituto Israelita de Ensino e Pesquisa Albert Einstein (IIEPAE). São Paulo , SP, Brazil
• City of the host institution: São Paulo
• Associated researchers: Ana Cláudia Camargo Miranda ; Daniel Alexandre Baptista Bonifacio ; Denise Mayumi Tanaka ; Guilherme Alberto Sousa Ribeiro ; LILIAN YURI ITAYA YAMAGA ; Luciana Malavolta Quaglio ; Luciano Fonseca Lemos de Oliveira ; Marcus Vinicius Simões ; Marycel Rosa Felisa figols de Barboza ; Orfa Yineth Galvis Alonso ; Paulo Victor Dos Santos ; Solange Amorim Nogueira ; Wesley Pacheco Calixto
• Associated scholarship(s): 25/12560-8 - Monte Carlo Simulaltion Modeling and Development of Tools for Processing and Reconstruction of Tomographic Images from Projections Acquired with Pinhole Collimators., BP.TT

Abstract

In preclinical research, imaging instruments for single photon emission computed tomography (SPECT) allow to evaluate the biodistribution of gamma ray emitting radiopharmaceuticals inside the animal's body in vivo, in contrast with the ex vivo assessment of biodistribution, where the animal is euthanized to determine the radioactivity in individual organs in gamma counters. Additionally, this Molecular Imaging technique allows to evaluate longitudinal evolution of physiopathogenic processes, as well as the effect of experimental protocols, in the same animal. With that, several aspects as the statistical value of the results and the number of animals are improved, producing results in shorter times, with lower costs and with improved quality.In the last 15 years, animal-dedicated imaging instruments have been developed and made commercially available (e.g., Triumph, U-SPECT, Albira, nanoScan, Molecubes), although its high cost limits accessibility. As an alternative and more accessible strategy, some research teams have proposed to upgrade SPECT clinical gamma cameras to acquire images of research animals. In this case, it is necessary to develop the appropriate hardware and software tools to reach the aimed spatial resolution, most commonly using pinhole collimators, as well as to produce the tomographic image from the recorded projections. This alternative has as an important advantage that the user has the knowledge of technical and developmental details, which makes easier to correct technical problems or to include novel recording or analysis strategies, in agreement with the specific research objectives.During a previous project, performed under the FAPESP's sponsorship (FAPESP process 07/50339-3), our research group developed and evaluated an upgrading device for SPECT clinical gamma cameras, in order to obtain images of small research animals, tested in conjunction with Siemens Orbiter, DST-Sopha, Elscint and Discovery GE cameras. In that project, hardware and software details were studied, which were necessary to reach the proposed objective. In addition, the implemented solution was tested in the characterization of animal models of cardiac and neurologic diseases, by producing images of cardiac perfusion in hamsters and rats, and brain perfusion in rats, with participation in the publication of eight scientific articles and in the formation of M.Sc. and Ph.D. researchers.In this project, we propose to develop a new version of the upgrading device, with the aim of improving characteristics related to spatial resolution (£1 mm), acquisition and reconstruction times (<10 min and ~5 min) and usable radioisotope's emission energy (up to ~250 keV). New versions of the tomographic reconstruction algorithm will be studied, based on classical programming techniques as well as artificial intelligence strategies. As a method to improve acquisition time, the use of multipinhole collimators without image overlapping will be considered. We will also evaluate a shielding and collimator configuration that allow to record images in energies higher than the emission energy of 99mTc. The implemented solution will be evaluated producing tomographic images from Monte Carlo simulations and records from physical phantoms, as well as with biological targets in small research animals in vivo.Finally, the execution of this project will contribute with the formation of new researchers, while strengthen a research line in Instrumentation and Molecular Imaging in our institutions. In the process, a research platform for development and evaluation of radiopharmaceuticals will also be reinforced, which has already a long trajectory in our institution. In addition, following the open science guidelines, the developed tools, both hardware and software, will be made available for free use by the scientific community, as it has been done with the previously developed tool. (AU)