SUMMARY
We are proposing a new approach to a hybrid imaging modality that has been called “b+g” or “pamma-positron”
Imaging [Gri07] that promises to simultaneously overcome 1) the sensitivity limits of single-gamma-ray-photon
emission imaging, 2) the challenge of distinguishing between two different positron-emitting isotopes, and 3) the
physics-based spatial resolution limits inherent in radioisotope imaging based on detection of positron-
annihilation photons alone [Lan14]. The intent is to significantly advance molecular imaging of the human brain
by allowing visualization of smaller substructures, quantification of smaller amounts of radiotracer uptake, and
simultaneous measurement of multiple dynamic and spatial uptake patterns in advanced multi-isotope studies
of normal brain function. The required elements to make this feasible comprise i) a detector approach for
annihilation and gamma-ray photons that can yield rich data for precise energy, position, and timing estimation
for both photoelectric and Compton interactions, ii) processing algorithms in firmware and software to sort and
make optimal use of the various combinations of signals that can occur with and without coincidence, iii)
reconstruction algorithms based on likelihoods that incorporate probabilities of emission, detection, positron
range, non-collinearity, and Compton kinematics, and iv) detection and compensation for attenuation and subject
motion – effects that if not addressed will become limiting factors for resolution and image quality.
In contrast to early efforts to accomplish b+g imaging with liquid xenon detectors [Gri07], scattering
detectors as inserts into PET scanners [Yos20], or planar semiconductor detectors paired with scintillation
cameras [Lan14], we propose instead to develop and demonstrate a single detector technology and associated
data processing methods that can be used for both 511 keV annihilation photons and the higher-energy, singly-
emitted gamma rays. Abbaszadeh (MPI) and Levin have pioneered an edge-on crossed-strip cadmium zinc
telluride (CZT) detector approach to PET detectors that provides an ideal starting point [Abb16]. Among their
attributes are high stopping power based on the edge-on geometry, 3D positioning that minimizes parallax,
excellent energy resolution, and dynamic range up to 1.2 MeV in maximum photon energy deposited per
interaction. Furthermore, when a photon undergoes an initial scatter followed by a photoelectric absorption, these
modules yield data vectors that allow position and energy estimation for both interactions that can be analyzed
with Compton kinematics [Abb17].
We will carry out a 2-year simulation and proof-of-principle phase (UG3) in which we demonstrate b+g
detection with edge-on CZT modules and measure detector characteristics, develop simulations that support
reconstructions, and demonstrate acquisitions with single and multiple isotopes. We will carry out a three-year
UH3 phase to build a tomographic system with a field of view sufficient to investigate imaging of sophisticated
dynamic phantoms and in vivo imaging of rodent brain, as a design study and precursor to a human brain system.
