Project description
Two tracers for better brain imaging
Positron Emission Tomography (PET) is a medical imaging technique that exposes in vivo brain-wide metabolic processes. Before PET image acquisition, a tracer is injected into the patient. PET image acquisition is limited to one tracer, thus only displaying one metabolic process in the brain. 'Dual tracer' PET images give novel insights for new drugs for Alzheimer's and increase understanding of the brain. The EU-funded Dual Tracer PET project will use machine learning techniques to develop and implement a reconstruction algorithm for dual tracer PET resulting in two separate PET images with high image quality. The results will help in understanding neuronal communication in the brain.
Objective
Positron Emission Tomography (PET) is a medical imaging technique that displays in vivo brain-wide metabolic processes. Prior to PET image acquisition, a tracer, labelled with a positron emitting isotope, is injected to the patient. This so-called radiotracer distributes over the body and accumulates in e.g. inflammatory tissue.
To date, the acquisiton of a PET image is limited to one tracer, thereby only displaying one metabolic process in the brain. However, the acquisition of two separate PET images acquired with different tracer simultaneously would give novel and unique insights in the communication between e.g. neurotransmitter systems. These insights could be used for the development of new drugs for e.g Alzheimer patients and would increase the understanding of the healthy and diseased brain.
The acquisition of 'dual tracer' PET images can be achieved by combining two radiotracers with different properties: The first 'standard' radiotracer emits two photons that are recorded by the PET system, while the second 'non-standard' tracer emits two photons and an additional gamma ray. By identifying the additional gamma ray, the photon detections of the two tracers can be separated and two PET images can be reconstructed.
'Dual tracer' PET image reconstruction yields challenges due to undesirable effect of photon detection. In this project, a machine learning algorithm will be trained to detect and correct for these effects. Moreover, the image acquired with the 'non-standard' tracer yields low image quality. A Convolutional Neural Network will be trained to denoise this PET image.
In the proposed project, a reconstruction algorithm for dual tracer PET based on machine learning techniques and resulting in two separate PET images with high image quality will be developed and implemented. The proposed project will demonstrate if dual tracer PET imaging has future potential and will help to understand the highly interactive neuronal communication in the brain.
Fields of science
Not validated
Not validated
- medical and health sciencesclinical medicineradiologynuclear medicine
- engineering and technologymedical engineeringdiagnostic imaging
- natural sciencescomputer and information sciencesartificial intelligencemachine learning
- natural sciencesphysical sciencestheoretical physicsparticle physicsphotons
- natural sciencescomputer and information sciencesartificial intelligencecomputational intelligence
Programme(s)
- HORIZON.1.2 - Marie Skłodowska-Curie Actions (MSCA) Main Programme
Funding Scheme
HORIZON-TMA-MSCA-PF-EF - HORIZON TMA MSCA Postdoctoral Fellowships - European FellowshipsCoordinator
52428 Julich
Germany