Opendata, web and dolomites

Report

Teaser, summary, work performed and final results

Periodic Reporting for period 2 - iMPACT (innovative Medical Protons Achromatic Calorimeter and Tracker)

Teaser

Hadron-therapy is a leading-edge technique which exploits the particular energy deposition prole (Bragg peak) protons or heavy-ions exhibit to target and destroy tumors within the human body. The beneficial aspects of this technique over other, more established, procedures...

Summary

Hadron-therapy is a leading-edge technique which exploits the particular energy deposition prole (Bragg peak) protons or heavy-ions exhibit to target and destroy tumors within the human body. The beneficial aspects of this technique over other, more established, procedures, such as X-ray therapy, have been extensively reported [1]. The effectiveness of a hadron-therapy treatment is strictly related to the accuracy of the knowledge of the tissues density or, equivalently, stopping power (SP) distribution: an accurate 3D map of the body SP makes it possible to precisely determine the position of the Bragg peak as a function of the beam energy. However, the effectiveness of the hadron-therapy procedure is currently limited by the necessity to rely on body density maps produced with X-ray Computed Tomography (X-ray CT), which cannot deliver maps accurate enough to fully exploit the intrinsic accuracy of the technique [2]. This is mainly due to the different behavior inside matter of X-ray and hadrons. A Computed Tomography performed with protons (pCT), instead of X-ray, would therefore improve the accuracy of the SP maps and lead to an enhancement of the treatment effectiveness, as the particles used for both the imaging process and the treatment present the same energy-loss behaviour. Recent studies (U. Schneider et al. [3]) confirmed that pCT can potentially be a factor 2:5 better, with respect to X-ray CT, in terms of SP values accuracy, with, at the same time, at least a factor 50 lower deposited dose. However the spatial resolution is expected to be worse for pCT than X-ray CT, mostly due to protons multiple Coulomb scattering (MCS) [4].

The iMPACT Project, innovative Medical Proton Achromatic Calorimeter and Tracker, aims to design and develop a pCT scanner, with the ultimate goal of demonstrating the viability of the pCT technique in a realistic clinical environment. A pCT scanner is composed by a tracker and a calorimeter. The tracker must provide the particles position and angle before and after they pass through the target (the human body), while the calorimeter has to measure the residual energy of the passing particles (Figure 1). By combining the track and energy information of about one billion protons passing through the target from different directions, it is possible reconstruct a 3D image of the target itself.

The goal of the iMPACT project is therefore to build a proton scanner fast enough to take a full 3D image of the targeted body part (by recording about 1 billion protons passing through it) in few seconds, so that the patient would not move, or even breath, during the irradiation. This would demonstrate the feasibility of a clinically viable pCT system. A successful outcome would mean improving the quality of cancer heavy-ions treatment, as well has creating a 3D body imaging tool less harmfull than current CT scanner, due to the lower dose released to record an image.


[1] Suit H. et al., Proton vs carbon ion beams in the denitive radiation treatment of cancer patiens, Radiother. Oncol., 95 (2010) 3-22.
[2] Sadrozinski H.F.W. et al., Toward proton computed tomography. IEEE Trans. Nucl. Sci., 51 (2004) 3-9.
[3] Schneider U.et al., First proton radiography of an animal patient. Med. Phys., 31 (2004) 1046-1051.
[4] Zygmanski P. et al., The measurement of proton stopping power using proton-cone-beam computed tomography, Phys. Med. Biol., 45 (2000) 511.

Work performed

- pCT
One of the drawbacks of pCT, when compared to a standard X-ray CT, is that protons undergo Multiple-Coulomb scattering (MS) as they move through matter. This affects the image spatial resolution (limited to about 5 mm as long as a standard tomography approach is employed [1]), blurring the resulting 3D reconstructed image. This effect can be overcome by adding the knowledge of each individual proton trajectory. A pCT scanner is therefore usually composed by two tracking stations, one before and one after the patient, to measure the proton entry and exit trajectories, and by a calorimeter, to measure the residual proton energy (Figure 1). The number of required proton tracks to reconstruct a 3D image with the necessary resolution (spatial resolution better than 1 mm, energy resolution better than 1%) has been estimated in [2]: in a volume of 10 x 10 x 10 cm3, at least 10^8 tracks have to be measured. Present state-of-the-art pCT prototypes made excellent advancements in the field, but they still show some limitations for a practical implementation. The custom-made technologies typically used (micro-strips [3][14], scintillating fibers [5], gas detectors [6]) limit their convenient application in commercial systems or they are undesirable in a hospital environment (due to the use of gas detectors and/or high voltages). In addition, the above mentioned pCT prototypes are limited in terms of data acquisition speed, so that they all require exposure times of the order of many minutes, too long for a practical implementation of the technique [7] due to the inevitable patient movements of the patient (breathing) and discomfort over such long period.

- Tracker development

First reporting period
The guidelines for the tracker are to develop a fast sensor, capable of handling a particle flux of 100 kHz cm^-2, with a low material budget to minimize multiple scattering from the protons. The sensor will be realized with commercially available CMOS processes, to reduce costs and ensure scalability. To simplify development and commissioning, the tracker will be organized as a modular composition of independent sensors. To fulfill the requirements of a real clinical medical application, the active surface must exceed 10 x 10 cm^2. To realize the tracker sensor, the iMPACT team is collaborating with the ALICE experiment at CERN. The starting point is the ALPIDE pixel sensor [8] that the ALICE collaboration has developed for the upgrade of the Inner Tracking System (Figure 2).
The ALPIDE sensor has been tested in July 2017 at TIFPA beam line hosted at the ATREP Proton Therapy Centre (Trento, Italy) with protons of energies between 70 MeV and 228 MeV. Figure 3 shows a quick demonstrative radiography of a plastic ball pen with the ALPIDE sensor and 70 MeV protons: it clearly illustrates the potential of large area MAPS sensors in medical imaging applications. Material with different densities can be distinguished: the plastic tip, the iron spring, the ink straw and the steel ballpoint (the densest and consequently the darkest part). Note that what in figure is NOT a classical image, i.e. the integral of the particle flux for each pixel, but a reconstructed map where every single particle (70 MeV proton in this case) has been distinguished and its coordinates and time recorded.
Another experimental pixel sensor, the OrthoPix sensor, has been fabricated and successfully tested (Figure 4). While much smaller than the ALPIDE, the OrthoPix sensor employs a very innovative readout architecture, specifically developed for the iMPACT project, and has been used as test-bed for many novel features and solutions. These advancements will be integrated together with the excellent pixel design of the ALPIDE sensor in the final iMPACT sensor, foreseen to be realized for the end of the iMPACT project.

Second reporting period
Based on the results reported at the end of the first period, it was established to proceed with the development of an improved se

Final results

\"- Sensor
We got the first proton image using a Monolithic Active Pixel Sensor of large area (30 x 15 mm^2), see Figure 3. Note that what in figure is NOT a classical image, i.e. the integral of the particle flux for each pixel, but a reconstructed map where every single particle (70 MeV proton in this case) has been distinguished and its coordinates and time recorded. This is the kind of information that, coupled with the energy measured in the calorimeter, allows reconstruction the 3D structure of the target. With the present configuration we can record up to 2x10^5 particles cm^-2 s^-1, which is equivalent to 2x10^8 particles cm^-2 s^-1 over an area of 10x10 cm^2 during 10s integration time. With the improved sensor version (currently in development phase) we plan to improve this speed by at least a factor 10, therefore reaching the target of more than 10^9 particles tracked in less than 10s over the target area.
The present sensor demonstrates that the sensing diode is effective in operating with proton of medical energy, and now the effort is on improving the readout architecture, tailoring it to further improve speed.

- Calorimeter
The early prototype of the iMPACT calorimeter has been extensively tested. The success of tests demonstrates that a full calorimeter is indeed feasible, and validates the idea of an hybrid energy-range calorimeter as an effective component of a fast and accurate proton tomography scanner. The prototypes also demonstrated that a global tracking speed of about 1 GHz (10 MHz cm^-2) is achievable, and that the energy resolution is actually around 1%, ultimately limited by the protons straggling itself. The calorimeter is completely modular and low cost (it uses only commercially available parts and components) when compared to present state of the art system, making it possible to employ it in other applications where high speed and resolution are necessary.

The construction of the final version of the calorimeter has started. The design uses group of 32 finger scintillators (Fig. 5) read out by SiPM and a front-end analog electronics. Each group of 32 fingers is connected to an FPGA, which stores locally the data flow in a DDR ram during the acquisition phase. The time resolution of the calorimeter is 4 ns (250 MHz) in the present implementation, but could be extended toward 2.5 ns if needed. Each signal generated by a passing protong is recorded as a 4 bits value for the amplitude (to distinguish between protons in the plateau or in the Bragg peak), a geographical identifier of the finger position, and a 32 bit long timestamp to allow linking the data to the tracker output in the final database. The first building blocks are currently under testing with cosmic rays (Fig. 12), and will be soon tested at a medical beam at the TIFPA facility in trento, likely by the end of the spring. Then a production of 8 to 12 modules is foreseen to complete the calorimeter prototype.

- Expected results
So far the project is on track, with the calorimeter prototypes having being validated in advance respect to the schedule, and the sensor design being slightly behind the schedule. By adopting the \"\"plan b\"\" solution of using the ALPIDE sensor it was possible to keep the tracker development on track, and at the same time to start the new sensor design with better know-how of the final specifications. We are therefore confident that the project will hit its initial goals of building a scanner prototype which will demonstrate the proton CT is a feasible and practical technique, i.e. applicable in a real medical environment and economically viable.Next steps foreseen in the project are the following:
1) Completion of the full calorimeter by the spring of 2019.
3) Design of a new sensor will start mid 2018, first submisison foreseen autumn 2019, tape-out early 2020.
4) Completion of the full tracker by summer 2019.
5) Testing of the full calorimeter and tracker by the end of 2019.
6) Completion of the new sensor duri\"