Abstract:Dual-tracer positron emission tomography (PET) can characterize various aspects of physiology and function by a single scan, both of which increase the accuracy of the diagnosis and benefit the patients significantly. However, the separation of dual-tracer PET images can be challenging as they yield indistinguishable 511-keV photon pairs. The existing methods rely on the differences in the kinetic behaviors of the compartment models and radioactive decay to differentiate dual tracers with staggered injection, thereby limiting the level of separation precision and increasing the dual-tracer scanning time. We herein propose a new method to separate the mixed dual-tracer signals with equal half-lives that allow the simultaneous injection of the two tracers. A deep belief network (DBN) framework with a supervised mechanism was constructed to map the complex relationship between dual-tracer signals and two individual signals. The DBN is composed of three restricted Boltzmnan machines (RBMs) and a supervised neural network with the same structure. Correspondingly, the whole training process includes two primary steps: 1) a forward pretraining by three RBMs and 2) a fine-tuning by the supervised neural network with the back-propagation (BP) algorithm. We validated the proposed DBN framework on four dual-tracer groups using the Monte Carlo simulation with different counts levels: 1) [F-18] fluorodeoxyglucose-[F-18] fluorothymidine on the Hoffman brain phantom; 2) [Cu-62] ATSM-[Cu-62] PTSM on the Zubal thorax phantom; 3) [C-11] flumazenil (FMZ)-[C-11] dihydrotetrabenazine (DTBZ) on the complex brain phantom; and 4) [C-11] FMZ-[C-11] DTBZ on the six-ball phantom. We demonstrate that our method can provide a satisfactory performance regarding the separation of dual tracers labeled by the same radionuclide with the order of 10(6) to 10(7) total counts.

Simple deep-learning approach for alpha-decay half-life studies

Simple deep-learning approach for -decay half-life studies

Simple deep-learning approach for <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"><mml:mi>α</mml:mi></mml:math> -decay half-life studies

Deep learning on nuclear mass and $α$ decay half-lives

Deep learning approach to nuclear masses and -decay half-lives

Bayesian Optimization Approach to Model-Based Description of Α Decay

$\Beta$-Decay Half-Lives Studied Using Neural Network Method

Unified description of decay and cluster radioactivity using the neural network approach and universal decay law

Calculating -decay half-lives with artificial neural networks considering the effects of angular momentum and deformation

Study α decay and proton emission based on data-driven symbolic regression

Deep-Learning-Based Separation of a Mixture of Dual-Tracer Single-Acquisition PET Signals with Equal Half-Lives: A Simulation Study

Diffuseness effect and radial basis function network for optimizing decay calculations

Deep Learning for Nuclear Physics

Α-Decay Half-Lives of Ground and Isomeric States Calculated in a Unified Theoretical Framework

Deep Learning-based Alpha Particles Spectroscopy with Solid-state Nuclear Track Detector CR-39

Comparative Study of Neural Network and Model Averaging Methods in Nuclear Β-Decay Half-Life Predictions

Half-lives of Α Decay from Natural Nuclides and from Superheavy Elements

Competition Betweenαdecay and Cluster Radioactivity for Superheavy Nuclei with a Universal Decay-Law Formula

Predictions of nuclear $β$-decay half-lives with machine learning and their impacts on $r$ process

Theoretical Predictions on Α-Decay Properties of Some Unknown Neutron-Deficient Actinide Nuclei Using Machine Learning

Noise Reduction of Nuclear Magnetic Resonance Spectroscopy Using Lightweight Deep Neural Network