• PET
• hadrontherapy

The Positron Emission Tomography is a technique of medical imaging
widely used to diagnose tumoural masses making use of the capability of a
radiotracer (18F − F DG) to reach, thanks to biological metabolism, sick
organs; 18 F is a β+ emitter with half-life of about 109 minutes, after the
decay the positron annihilates with an electron of the body emitting two
gamma rays back-to-back of energy of 511 keV. Then, it is possible to go
back to the position of the annihilation process by revealing the two photons
with detectors placed all around the patient.
The use of particles beams (like: photons, electrons and hadrons) for
medical purpose has been developed thanks to progress in the construction
of particle accelerators. In particular hadrontherapy exploits heavy particle
beams to destroy tumoural masses inside the body. Particles commonly used
are protons and carbon ions, they deliver almost all their energy in a narrow
peak, whose depth depends upon the energy of the ions, so it is possible to
reach deep organs by tuning the energy of the beam. Another advantage to
make use of hadrons comes from the fact that the energy released to healthy
tissues is less than that released by photons or electrons, considering the
same dose deposited on tumor.
In the last decades it began to think to the possibility of applying the
principle of photon detection of a PET scanner to monitor the dose released
in tissues by beams employed in hadrontherapy.
This is feasible because when an hadron beam crosses matter it produces
isotopes, like 11 C or 15 O, that are β+ emitters; then, after radioactive decay
and the annihilation with an electron, they will produce two γ rays, which,
once revealed, give the distribution of isotopes and so it is possible to localize
where the beam has interacted.
In this work of thesis it has been characterized a PET scanner, named
DoPET, planned for the monitoring of the dose in hadrontherapy. It consists
of two planar heads of detector, each of them is composed by four independent
scintillator LYSO matrices, coupled with positions sensitive photomultiplier
tubes and connected with an electronic system of data acquisition.
The analysis pertains to the performance of the device, and it covers
various aspects of the system. The measures were made referring to the
NEMA standard protocol, but, since it refers to the PET scanners with a
circular geometry used for research on small animals, some of these analysis
had to be adapted to planar geometry.
The first study is linked with the calibration of the system, which is necessary before the reconstruction of images. The first step of this calibration
is the pixel identification, it is required because on the edges of the crystal
pixels are not always well resolved, then it is necessary the energy calibration
to convert spectra from raw ADC channels to energy channels.
It was noticed that positions of pixels changed, respect to those identified
with calibration, with varying of the rate of the source. So it is tested a
method to correlate the correction of pixel maps with the rate of the source
and then it is included in the reconstruction software to avoid the wrong
attribution of events in pixels changing the activity.
The analysis about the dead-time of the device concerns the estimation
of the Constant Fraction Discriminator dead-time and the verification that
modularity of the detector improves dead-time performances of the device, as
predicted by simulations. Furthermore the dead-time correction is introduced
inside the reconstruction software to obtain the correct rate of the source.
The study about the spatial resolution of the device was done with two
different radioactive sources, and it was observed the same behavior by the
system, so that the resolution at the center of the device was worse than in
other points of the Field Of View, this fact is due to parallax error, which is
more significant at the center rather than on borders, because of the geometry
of the scanner.
The study about the Noise Equivalent Count rate is important to know
the maximum range of applicability in-beam, namely at which rate the signal
to noise ratio is still favorable; in fact increasing the activity of the source
noise grows more than signal, degrading the quality of data. The NEC curve
is calculated as the ratio between the true coincidence rate square and the
sum of all acquired coincidences; from the position of the peak of the curve
it is possible to go back to the maximum advantageous value of activity.