Voxel Imaging PET Pathfinder (VIP)

VIP Pathfinder is an ambitious project with the aim to prove the feasibility of constructing a PET/Gamma scanner from pixel solid-state-detector (CdTe).
Since this is a “Pathfinder” proposal, it will be enough to prove the concept by constructing two sectors of PET ring using pixel CdTe detector. With these two sectors one can study the energy resolution, the spatial resolution, the detection efficiency, the overall image resolution versus dose, and finally to check its performance under high magnetic field.

VIP project is being funded by FP7-ERC-AG. It has started in July 2010 and for a period of 48 months.

Positron-Emission-Tomography (PET) scanners play an important role in cancer diagnosis. Their accuracy overtakes that of the conventional diagnostics systems. They are used, alone or in combination with other imaging systems, such as CT or MRI, as follow up to the therapy treatment in which the PET shows the rate at which the tumor size has been reduced, for example. This in turn can give an indication to the doctor to tune the treatment for optimum results. Usually the PET image gives the final clean health bill for patients that have undergone irradiation and chemotherapy treatments. When PET is used with other type of image modality such as MR the combined tool PET-MR is transformed into a powerful tool for molecular imaging and in particular for brain imaging.

The current best detectors for PET are based on LSO crystals that are usually made of 4mm x 4mm x 10mm coupled to PMT, APD, or similar photon-sensitive device. The light yield of LSO for 511keV gamma is about 4000 phe. At best, the FWHM that can be achieved with LSO at 511keV is around 10%. This intrinsic limitation in the energy resolution restricts the ability to remove scattered events which are a significant noise contamination to the reconstructed image. The typical length (in the radial direction) of the LSO crystals is about 10mm, which implies a significant uncertainty of the impact point in the radial direction and induces an error in the projection of the Line of Response (LOR; the line that connects the supposedly back-to-back events). This error deteriorates the quality of the reconstructed image.


 Figure 1. A simplified drawing showing how the PET image data is acquired.

Figure 2. On the right-hand-side there is a typical detector module for a PET scanner based on Scintillator Crystals. On the left-hand-side, there is a module detector for the VIP PET scanner. This detector module has 8000 voxel and each has its own pre-amp, shaper, and 10 bits ADC. 

In “Voxel Imaging PET Pathfinder” (VIP), we develop a novel detector design which uses a pixelated room temperature solid-state-detector (CdTe) coupled to readout front end electronics, instead of scintillating crystals coupled to a photon-sensitive sensor.
With VIP detector one can achieve the following:
* The detector can be segment into millimeter-size pixels (or voxels), something that it is not possible to achieve with scintillating crystals.
* The CdTe detector can achieve a FWHM energy resolution of less than 1% for 511keV photons. This allows us to eliminate most of the scattered events and thus keep the golden events.
* The VIP design allows an adequate depth (4cm) of absorption using CdTe detectors, to achieve high detection efficiency for 511keV photon, something that it is not possible to achieve with PET based on Crystal detectors.
* The modular-Unit-detector of Pixel-PET can be designed to have rectangular parallelepiped shape, instead of trapezoidal parallelepiped geometry, so that it can be arranged to form a flat-panel detector, instead of cylindrical, to be used as Gamma Camera as shown below
* The electric field inside the solid state detector is parallel to magnetic field of the MRI. Thus, the response of the solid state detector is not affected by the strenght of the magnetic field.
Due to the fact the detector module for the VIP PET scanner are trapezoidal parallelepiped is easier to form a perfect circle, increasing the detection efficiency.


 Figure 3. The right-hand-picture shows the cracks at the contact edges between two adjacent module detectors for PET scanner based on scintillating crystals. The left-hand-picture is the hermetic geometry provided by VIP PET scanner.


Figure 4. Trapezoidal-parallelepiped geometry for the pixel CdTe detector module.

Results from GAMOS (Geant4 simulation dedicated for medical applications) using Derenzo Phantom with VIP detector suggest that one can achieve the same image quality of the Clear-PET device but using 25 times less dose. This is possible because VIP has cutting edge quality (listed above) when it is compared to other PET scanners based on Scintillating Crystal detectors. In VIP we use only “golden events” to construct the PET image where the noise contribution to the image is reduced to the minimum.

Figure 5. Image reconstructed of Derenzo Phantom, based on GAMOS simulation for Clear-PET and VIP respectively.
The dose used for the reconstruction image for Clear-PET is 25 times the dose used for VIP.

VIP scanner with CdTe

Crystal PET

Figure 6. VIP scanner with CdTe vs. Crystal PET
The ratio of scattered vs true events in VIP scanner using CdTe is reduced to approximately 3 %, compared to 90% for Crystal PET. Moreover, the event rate in VIP, which reflects the detection efficiency factor, increased to 6.1 counts/kBq, compared to 2.5 counts/kBq for Crystal PET.

Though the VIP project is focusing on PET scanner, VIP concept can be extended to be used for Gamma Camera, SPECT, and Compton Camera. In fact it is ideal for Compton Camera because of two basic parameters:
* Its excellent energy resolution of the deposited energy with FWHM better than 1%
* Its excellent spatial resolution due to its nature as 3D detector that eliminates the parallax effect


Figure 7. A possible configuration for Gamma ray detector.

The VIP project will construct 2 supper-module-detectors, as shown below, in which they are separated from by 40 cm, a typical gantry of head-PET. Derenzo phantom source will be placed on a sliding/rotating axis that will allow it to be positioned at different Z value from the centre of the setup. From this setup we will be able to obtain a close to reality case for the performance of a full VIP scanner



Figure 8. 2 supper-module-detector.  


We use own and third-party cookies to improve our services by analyzing your browsing habits. If you continue browsing, we will consider that you allow us to use them. You can change the settings or get more information on our "Cookies policy".

I accept cookies from this site.