Instrumentation for microSPECT and microPET imaging

Division Of Basic Sciences - Nci

Investigators

Linked publications, trials & patents

Abstract

The physics group within the Molecular Imaging Branch continues to advance the state of knowledge in camera design and specific instrumentation that aids research and may also be important for clinical translational imaging. Specifically, the major project of the physics group is the design of new instrumentation. Currently, PET cameras rely on a process called iterative reconstruction to generate PET images. This is a wasteful method because it introduces considerable noise into the image. A superior way to generate PET images would be to develop Time of Flight (TOF) imaging in which the location of the annihilation event could be exactly located by the relative time the two emitted photons take to reach their respective detectors. If the time resolution of the detector is sufficient the annihilation event can be specified to within 3-4mm. This would be sufficient to generate images of equal resolution to todays scanner but with much higher sensitivity. This project is a multiyear effort to develop a practical sub 30ns detector. Our team has decades of experience in camera and detector design and this project could have significant impact on the field. Additionally, we have an important secondary project. The problem this second project seeks to solve is one of inefficiency of Single Photon Emission Computed Tomography (SPECT). SPECT relies on a single emission (as opposed to dual emissions for PET) of a gamma ray. In order to localize where the event occurred that gave rise to the single photon emission considerable collimation in the form of tungsten barriers are installed over the detectors to ensure that only photons from a particular direction are allowed to reach the detector. This simplifies the formation of the image however, it is massively inefficient and <1% of the photons emitted are captured on a SPECT image. This results in relatively high amounts of radioactivity that need to be administered for a SPECT image to be generated. Our team is developing a "collimator-less" SPECT camera in which the collimators are removed but AI is used to reconstruct the source of the photon according the angle and depth of interaction with the crystal. Preliminary data suggests this camera architecture is feasible and results in excellent simulated images based on AI enhancements. This would increase the sensitivity of SPECT by at least a factor of 10 reducing the needed dose for SPECT imaging. Additionally, customized projects are underway to develop needed tools for molecular imaging. Alpha irradiation is difficult to detect. Using detectors also used at major nuclear accelerators (e.g. CERN in Geneva) we have developed a detector that can detect both alpha emissions and PET emissions on the same tissue enabling colocalization studies. In summary the Instrumentation group within MIB continues to accomplish major research milestones in developing the next generation of nuclear medicine cameras that will be more sensitive and thus, require less radioactivity than our current cameras.