Hugo Simoes

Orthogonal ray imaging with megavoltage beams: Simulated results with an anthropomorphic phantom

Orthogonal ray imaging is a new technique under investigation by our groups. It shows good potential for (1) dose verification in external beam radiotherapy and (2) very-low-dose computed tomography (CT-like) imaging. We have simulated with Geant4 the performance of four flat-panel-like perfect detectors for evaluating the capabilities of orthogonal ray imaging and portal imaging for assisting external beam therapy. The four detectors were positioned surrounding the head of the patient, three parallel and one perpendicular to the beam axis. Each detector covers an area of 185 mm × 185 mm and the simulation scores every particle seen by every detector. This allows for a second-stage investigation of optimum acceptance angles and energy thresholds, presented here. For demonstration purposes one small rectangular, sub-therapeutic beam with a front area of 20 mm × 5 mm, maximum entrance dose in the buildup region of 1.3 mGy, and tumor dose of 0.8 mGy, was shot at the region of the pituitary gland of an anthropomorphic phantom. Despite the low dose, visual inspection shows a remarkable agreement both with the predicted dose and with patient bone structures, collected with the orthogonal ray detectors. The portal imaging detector could not provide comparable information based on a single shot. In addition, seven small rectangular beamlets irradiating the region of the pituitary gland of the phantom with 4.6 mGy and simulating an intensity modulated radiation therapy treatment-like scenario were also analyzed, showing equally a visual agreement with the planed dose distribution. We finally evaluate the possibility of using rotation-free, scanned megavoltage orthogonal ray imaging (multi-slice collimation) for patient morphologic imaging. The technique provides images of high visual correlation with phantom morphology, therefore being of potential usefulness e.g. for low-dose, on-board patient imaging prior to a radiotherapy treatment, among other applications.

Rotation-free computed tomography with orthogonal ray imaging: First millimetric experimental results

Orthogonal ray imaging is a new technique under investigation by our groups. It shows good potential for (1) dose verification in external beam radiotherapy, and (2) very-low-dose computed tomography (CT-like) imaging, the latter being investigated here. We present simulated (Geant4) and experimental orthogonal ray imaging results of the visualization of a heterogeneous phantom irradiated from both sides with a 6-MV photon beam. The phantom was constructed with azimuthal symmetry in respect to the incoming beam directions. This imaging technique, named orthoCT, does not require rotational irradiation of the target. It is based on the detection of photons emitted at almost right angles in respect to the incoming photon flux, the latter being shot at the target only from two diametrically opposed directions. In addition to showing the visual correlation between the detected photon profile and the phantom density profile, we also show by simulation and experiment that a 2-mm phantom displacement is clearly distinguishable with orthoCT. A new, improved collimator is under construction aiming at achieving sub-millimetric accuracy in phantom positioning.

Preliminary characterization of the external proton beam from a PET cyclotron for use in neutron and proton radiobiology and other dosimetric studies

Different cyclotron models capable of accelerating protons up to 20MeV have been worldwide installed. Although their purpose is mainly positron emission tomography (PET) radioisotope production, they are equipped with several beam lines suitable for scientific research. Each beam line may typically deliver proton currents up to 150 μA (1×1015 particles/s). Radiobiological and dosimetric studies can be performed using these beam lines, which may contribute to further improve ion therapy and material radiation hardness results, among other applications. We report experimental results aiming at characterizing the proton beam achievable outside a PET cyclotron. In addition, we simulate this experimental setup by using Geant4. We show that simulation is consistent with previous published experimental data and with our first-measured results. These point to a beam angular spreading which may be utilized to establish an irradiation setup within the bunker. We estimate that the dose achievable with such setup may span 4-orders-of-magnitude, useful in radiobiology, ranging from 10mGy to 100 Gy. Finally, we show by simulation that neutron and γ-ray dose on a realistic, in-bunker target is negligible down to at most the 1% level. Further quantification for lower levels is ongoing.

Dose-free monitoring of radiotherapy treatments with scattered photons: Concept and simulation study

Modern radiotherapy (RT) techniques provide increasingly higher conformality, a potential invaluable clinical benefit to the patient. Consequently, in both single and multi-fractionated RT, patient misalignments and changing internal anatomy are also becoming more critical since higher conformality may equally represent a higher risk of target underdosage or organ-at-risk overdosage. Even with rigid fixation devices, maximum positioning errors higher than 1 cm are observable. In addition, anatomical morphological variations induced by cardiorespiratory or bowel motion, or RT-related biological responses, have been reported. The latter include tissue swelling, edema, inflammation, tumor shrinkage/growth, or filling of body cavities with unaccounted mucus or edematous tissue. State-of-the-art image-guided radiotherapy (IGRT) aims at providing feedback to the radiation oncologist in regard to these matters, some times at the cost of increased dosage (e.g. kilo and megavoltage IGRT), other times providing insufficient clinical information. We investigate a novel imaging system specially designed for monitoring both conventional and intensity/volumetric modulated photon radiotherapy (IMRT/VMRT, static and dynamic). The proof-of-principle and feasibility of such system indicate its potential for monitoring each field (and segment, for IMRT/VMRT) during all treatment fractions without whatsoever additional dose. We present a clear 2D correlation between the dose delivered in a heterogeneous phantom and the number of scattered photons detected perpendicular to the photon beam. Simulations of high-energy, multi-hole collimators show real-time, high-detectability of abnormal (though possible) irradiation scenarios with pertinent target morphological alterations, such as tumor dislocation or formation of edematous tissue.

On-line measurements of proton beam current from a PET cyclotron using a thin aluminum foil

The number of cyclotrons capable of accelerating protons to about 20 MeV is increasing throughout the world. Originally aiming at the production of positron emission tomography (PET) radionuclides, some of these facilities are equipped with several beam lines suitable for scientific research. Radiobiology, radiophysiology, and other dosimetric studies can be performed using these beam lines. In this work, we measured the Bragg peak of the protons from a PET cyclotron using a stacked target consisting of several aluminum foils interleaved with polyethylene sheets, readout by in-house made transimpedance electronics. The measured Bragg peak is consistent with simulations performed using the SRIM/TRIM simulation toolkit. Furthermore, we report on experimental results aiming at measuring proton beam currents down to 10 pA using a thin aluminum foil (20-μm-thick). The aluminum was chosen for this task because it is radiation hard, it has low density and low radiation activity, and finally because it is easily available at negligible cost. This method allows for calculating the dose delivered to a target during an irradiation with high efficiency, and with minimal proton energy loss and scattering.

Towards a high-dynamic dose-range irradiation setup for radiobioloy and radiophysiology

The number of cyclotrons capable of accelerating protons to about 20 MeV is increasing throughout the world. In Portugal, an IBA (Ion Beam Applications, Belgium) model Cyclone 18/9 cyclotron was installed at ICNAS (Instituto de Ciencias Nucleares Aplicadas a Saude) for positron emission tomography (PET) in 2010. Such facility is equipped with eight beam lines suitable for scientific research. Each beam line may deliver proton currents up to 150 μA (1×1015 particles/s). Radiobiological and dosimetric studies, among others, can be performed using these beam lines. In this work, we report on experimental results and Geant4 and SRIM/TRIM simulations, which aim at investigating the possible use of the 18 -MeV proton beam from the PET cyclotron to irradiate a selected region of a target with 1mm diameter. We prove by simulation and experiment that we are indeed able to irradiate a selected region of a target with 1mm diameter (i.e., the beam spot is very sharp). In addition, simulations indicate that neutron and γ-ray dose on a realistic target is negligible down to at most the 1 % level when compared with the proton dose. Moreover, the experimental and simulation results show good beam uniformity in the target-selected region. In addition to the previously-mentioned results, we measured the Bragg peak of the protons from ICNAS cyclotron by using a stacked target consisting of aluminum foils interleaved with polyethylene sheets, readout by custom-made electronics. Finally, calculations show that the proposed irradiation setup may span 4 orders of magnitude, ranging from 1 cGy to 100 Gy. Such a setup could satisfy user requirements in several fields, including ion therapy radiobiological and animal studies and biological science.

Experimental characterization of megavoltage beams for orthogonal ray imaging

OrthoCT (orthogonal computed tomography) is a potential new imaging technique that aims to acquire images of the volume to be irradiated immediately before or during a radiotherapy treatment. It potentially provides imaging with very low to eventually null dose, allowing to check if the morphology/anatomy of the patient and tumour are in agreement with the planned one. This technique relies on the detection of photons that are scattered in the patient and are emitted perpendicularly to the incident beam direction. To acquire the OrthoCT morphological images the scanning of the volume to be irradiated is done using pencil-like mega-voltage beams. The corresponding scan profile requires: (1) high homogeneity, so that variations can be associated only to dose/morphological alterations, and (2) high velocity, which favors multi-leaf collimator-based scans in respect to jaw-based ones. To compare the variability of a homogeneous beam with the variability of a scanned profile two scans with a cross-section of 5mm × 5mm (MLC-collimated) and 6mm × 6mm (jaw-collimated) were experimentally evaluated. The transverse profiles obtained with MLC-collimation in this work reveal a homogeneity with an intensity variability inferior to 1%, thus supporting OrthoCT imaging with morphology/anatomy changes superior to that value.

Impact of tumor contrast in orthogonal ray imaging: A prostate irradiation study

A new very-low-dose imaging technology to assist external-beam radiotherapy treatments has been proposed. This technique, called orthogonal ray imaging, does not require X-ray source rotation around the target. It is based on the detection of photons escaping the target at almost right angles with respect to the incoming photon flux. Modern image-guided radiation therapy techniques allow an accurate positioning of the patient, consequently improving the treatment accuracy. However, some of these techniques, such as electronic portal imaging and mega-voltage cone-beam computed tomography provide poor resolution in soft tissues. Other techniques like kilovoltage cone-beam computed tomography, allow a good visualization of these type of tissues but at the cost of an increase of the dose delivered to the patient, consequently increasing the risk of possible side-effects on the surrounding healthy tissues. One way to enhance the contrast of the lesion region is by injecting a contrast agent. The aim of this study was to evaluate through simulation the benefit of iodinated-contrast agent (Ultravist 370) administration in cases of prostate cancer in OrthoCT using the Geant4 simulation toolkit and the anthropomorphic phantom NCAT. We conclude that (1) in all studied scenarios it was possible to obtain OrthoCT images with good visual agreement with the simulated dose as well as the phantom pelvic structures with a maximum dose of the order of 2mGy and (2) the administration of 100mL of Ultravist 370 is not enough to allow the visualization of the tumor or the normal prostate tissue.

Development of a PET cyclotron based irradiation setup for proton radiobiology

An out-of-yoke irradiation setup using the proton beam from a cyclotron that ordinary produces radioisotopes for positron emission tomography (PET) has been developed, characterized, calibrated and validated. The current from a 20 μm thick aluminum transmission foil is readout by home-made transimpedance electronics, providing online dose information. The main monitoring variables, delivered in real-time, include beam current, integrated charge and dose rate. Hence the dose and integrated current delivered at a given instant to an experimental setup can be computer-controlled with a shutter. In this work, we report on experimental results and Geant4 simulations of a setup which exploits for the first time the 18 MeV proton beam from a PET cyclotron to irradiate a selected region of a target using the developed irradiation system. By using this system, we are able to deliver a homogeneous beam on targets with 18 mm diameter, allowing to achieve the controlled irradiation of cell cultures located in biological multi-well dishes of 16 mm diameter. We found that the magnetic field applied inside the cyclotron plays a major role for achieving the referred to homogeneity. The quasi-Gaussian curve obtained by scanning the magnet current and measuring the corresponding dose rate must be measured before any irradiation procedure, with the shutter closed. At the optimum magnet current, which corresponds to the center of the Gaussian, a homogenous dose is observed over the whole target area. Making use of a rotating disk with a slit of 0.5 mm at a radius of 150 mm, we could measure dose rates on target ranging from 500 mGy/s down to 5 mGy/s. For validating the developed irradiation setup, several Gafchromic® EBT2 films were exposed to different values of dose. The absolute dose in the irradiated films were assessed in the 2D film dosimetry system of the Department of Radiotherapy of Coimbra University Hospital Center with a precision better than 2%. In the future, we plan to irradiate small animals, cell cultures, or other materials or samples.

Experimental demonstration of induction by means of a transcranial magnetic stimulator coil immersed in a conducting liquid

The full potentialities of deep-brain transcranial magnetic stimulation (TMS) are presently limited by the so-called surface discontinuity effect. This effect is responsible for a strong reduction of the capability of TMS systems to reach satisfactorily deep-brain regions with biostimulatory and/or bioinhibitory purposes. Consequently, a large number of neuropathologies that could potentially profit from TMS remain either unknown, or yield clinical trials with inconclusive results. Previous simulation studies have pointed to the fact that the surface discontinuity effect may be reduced very strongly when the coil and the patient head are immersed in a conducting liquid. In this work we provide experimental evidence of this fact. For that, we have constructed a TMS system capable of inducing quasi-monophasic currents in saline solution of up to 1.5 A/m2 delivered in less than 200μs. Such current density and temporization are those typically used for TMS. We present results measured in an NaCl solution with a conductivity of ~0.11 S/m (0.05%w/v), i.e. three times smaller than the brain conductivity. This saline solution was inserted in a cylindrical container with a diameter of 125mm, therefore representing tentatively a small brain. A surface-to-center induced current ratio of -81% was measured when the brain-like container was stimulated with the TMS system surrounded by air. When both the stimulating coil and the brain-like container where immersed in a saline solution with a conductivity of 5.5 S/m (3.33%w/v), the surface-to-center induced current ratio dropped only -24%, therefore confirming qualitative expectations from previous simulation work. This experimental confirmation opens new possibilities for deep-brain TMS in neurology.