Monte Carlo evaluation of high-gradient magnetically focused planar proton minibeams in a passive nozzle.

Objective. To investigate the potential of using a single quadrupole magnet with a high magnetic field gradient to create planar minibeams suitable for clinical applications of proton minibeam radiation therapy.Approach. We performed Monte Carlo simulations involving single quadrupole Halbach cylinders in a passively scattered nozzle in clinical use for proton therapy. Pencil beams produced by the nozzle of 10-15 mm initial diameters and particle range of âˆ¼10-20 cm in water were focused by magnets with field gradients of 225-350 T m-1and cylinder lengths of 80-110 mm to produce very narrow elongated (planar) beamlets. The corresponding dose distributions were scored in a water phantom. Composite minibeam dose distributions composed from three beamlets were created by laterally shifting copies of the single beamlet distribution to either side of a central beamlet. Modulated beamlets (with 18-30 mm nominal central SOBP) and corresponding composite dose distributions were created in a similar manner. Collimated minibeams were also compared with beams focused using one magnet/particle range combination.Main results. The focusing magnets produced planar beamlets with minimum lateral FWHM of ∼1.1-1.6 mm. Dose distributions composed from three unmodulated beamlets showed a high degree of proximal spatial fractionation and a homogeneous target dose. Maximal peak-to-valley dose ratios (PVDR) for the unmodulated beams ranged from 32 to 324, and composite modulated beam showed maximal PVDR ranging from 32 to 102 and SOBPs with good target dose coverage.Significance.Advantages of the high-gradient magnets include the ability to focus beams with phase space parameters that reflect beams in operation today, and post-waist particle divergence allowing larger beamlet separations and thus larger PVDR. Our results suggest that high gradient quadrupole magnets could be useful to focus beams of moderate emittance in clinical proton therapy.

Experimental validation of magnetically focused proton beams for radiosurgery.

We performed experiments using a triplet of quadrupole permanent magnets to focus protons and compared their dose distributions with unfocused collimated beams using energies and field sizes typically employed in proton radiosurgery. Experiments were performed in a clinical treatment room wherein small-diameter proton beams were focused by a magnet triplet placed immediately upstream of a water tank. The magnets consisted of segments of Sm2Co17 rare-earth permanent magnetic material adhered into Halbach cylinders with nominal field gradients of 100, 150, 200, and 250 T m-1. Unmodulated beams with initial diameters of 3 mm-20 mm were delivered using a single scattering system with nominal energies of 127 and 157 MeV (respective ranges of ~10 cm and 15 cm in water), commonly used for proton radiosurgery at our institution. For comparison, small-diameter unfocused collimated beams were similarly delivered. Transverse and depth dose distributions were measured using radiochromic film and a diode detector, respectively, and compared between the focused and unfocused beams (UNF). The focused beams produced low-eccentricity beam spots (defined by the 80% dose contour) at Bragg depth, with full width at 80% maximum dose values ranging from 3.8 to 7.6 mm. When initial focused beam diameters were larger than matching unfocused diameters (19 of 29 cases), the focused beams peak-to-entrance dose ratios were 13% to 73% larger than UNF. In addition, in 17 of these cases the efficiency of dose delivery to the target was 1.3× to 3.3× larger. Both peak-to-entrance dose ratios and efficiency tended to increase with initial beam diameter, while efficiency also tended to increase with magnet gradient. These experimental results are consistent with our previous Monte Carlo (MC) studies and suggest that a triplet of quadrupole Halbach cylinders could be clinically useful for irradiating small-field radiosurgical targets with fewer beams, lower entrance dose, and shorter treatment times.

Monte Carlo evaluation of magnetically focused proton beams for radiosurgery.

The purpose of this project is to investigate the advantages in dose distribution and delivery of proton beams focused by a triplet of quadrupole magnets in the context of potential radiosurgery treatments. Monte Carlo simulations were performed using various configurations of three quadrupole magnets located immediately upstream of a water phantom. Magnet parameters were selected to match what can be commercially manufactured as assemblies of rare-earth permanent magnetic materials. Focused unmodulated proton beams with a range of ~10 cm in water were target matched with passive collimated beams (the current beam delivery method for proton radiosurgery) and properties of transverse dose, depth dose and volumetric dose distributions were compared. Magnetically focused beams delivered beam spots of low eccentricity to Bragg peak depth with full widths at the 90% reference dose contour from ~2.5 to 5 mm. When focused initial beam diameters were larger than matching unfocused beams (10 of 11 cases) the focused beams showed 16%-83% larger peak-to-entrance dose ratios and 1.3 to 3.4-fold increases in dose delivery efficiency. Peak-to-entrance and efficiency benefits tended to increase with larger magnet gradients and larger initial diameter focused beams. Finally, it was observed that focusing tended to shift dose in the water phantom volume from the 80%-20% dose range to below 20% of reference dose, compared to unfocused beams. We conclude that focusing proton beams immediately upstream from tissue entry using permanent magnet assemblies can produce beams with larger peak-to-entrance dose ratios and increased dose delivery efficiencies. Such beams could potentially be used in the clinic to irradiate small-field radiosurgical targets with fewer beams, lower entrance dose and shorter treatment times.

Initial testing of a pixelated silicon detector prototype in proton therapy.

As technology continues to develop, external beam radiation therapy is being employed, with increased conformity, to treat smaller targets. As this occurs, the dosimetry methods and tools employed to quantify these fields for treatment also have to evolve to provide increased spatial resolution. The team at the University of Wollongong has developed a pixelated silicon detector prototype known as the dose magnifying glass (DMG) for real-time small-field metrology. This device has been tested in photon fields and IMRT. The purpose of this work was to conduct the initial performance tests with proton radiation, using beam energies and modulations typically associated with proton radiosurgery. Depth dose and lateral beam profiles were measured and compared with those collected using a PTW parallel-plate ionization chamber, a PTW proton-specific dosimetry diode, EBT3 Gafchromic film, and Monte Carlo simulations. Measurements of the depth dose profile yielded good agreement when compared with Monte Carlo, diode and ionization chamber. Bragg peak location was measured accurately by the DMG by scanning along the depth dose profile, and the relative response of the DMG at the center of modulation was within 2.5% of that for the PTW dosimetry diode for all energy and modulation combinations tested. Real-time beam profile measurements of a 5 mm 127 MeV proton beam also yielded FWHM and FW90 within ±1 channel (0.1 mm) of the Monte Carlo and EBT3 film data across all depths tested. The DMG tested here proved to be a useful device at measuring depth dose profiles in proton therapy with a stable response across the entire proton spread-out Bragg peak. In addition, the linear array of small sensitive volumes allowed for accurate point and high spatial resolution one-dimensional profile measurements of small radiation fields in real time to be completed with minimal impact from partial volume averaging.

Evaluation of the dosimetric properties of a diode detector for small field proton radiosurgery.

The small fields and sharp gradients typically encountered in proton radiosurgery require high spatial resolution dosimetric measurements, especially below 1-2 cm diameters. Radiochromic film provides high resolution, but requires postprocessing and special handling. Promising alternatives are diode detectors with small sensitive volumes (SV) that are capable of high resolution and real-time dose acquisition. In this study we evaluated the PTW PR60020 proton dosimetry diode using radiation fields and beam energies relevant to radiosurgery applications. Energies of 127 and 157 MeV (9.7 to 15 cm range) and initial diameters of 8, 10, 12, and 20mm were delivered using single-stage scattering and four modulations (0, 15, 30, and 60mm) to a water tank in our treatment room. Depth dose and beam profile data were compared with PTW Markus N23343 ionization chamber, EBT2 Gafchromic film, and Monte Carlo simulations. Transverse dose profiles were measured using the diode in "edge-on" orientation or EBT2 film. Diode response was linear with respect to dose, uniform with dose rate, and showed an orientation-dependent (i.e., beam parallel to, or perpendicular to, detector axis) response of less than 1%. Diodevs. Markus depth-dose profiles, as well as Markus relative dose ratio vs. simulated dose-weighted average lineal energy plots, suggest that any LET-dependent diode response is negligible from particle entrance up to the very distal portion of the SOBP for the energies tested. Finally, while not possible with the ionization chamber due to partial volume effects, accurate diode depth-dose measurements of 8, 10, and 12 mm diameter beams were obtained compared to Monte Carlo simulations. Because of the small SV that allows measurements without partial volume effects and the capability of submillimeter resolution (in edge-on orientation) that is crucial for small fields and high-dose gradients (e.g., penumbra, distal edge), as well as negligible LET dependence over nearly the full the SOBP, the PTW proton diode proved to be a useful high-resolution, real-time metrology device for small proton field radiation measurements such as would be encountered in radiosurgery applications.

Single-Plane Magnetically Focused Elongated Small Field Proton Beams.

We previously performed Monte Carlo simulations of magnetically focused proton beams shaped by a single quadrapole magnet and thereby created narrow elongated beams with superior dose delivery characteristics (compared to collimated beams) suitable for targets of similar geometry. The present study seeks to experimentally validate these simulations using a focusing magnet consisting of 24 segments of samarium cobalt permanent magnetic material adhered into a hollow cylinder. Proton beams with properties relevant to clinical radiosurgery applications were delivered through the magnet to a water tank containing a diode detector or radiochromic film. Dose profiles were analyzed and compared with analogous Monte Carlo simulations. The focused beams produced elongated beam spots with high elliptical symmetry, indicative of magnet quality. Experimental data showed good agreement with simulations, affirming the utility of Monte Carlo simulations as a tool to model the inherent complexity of a magnetic focusing system. Compared to target-matched unfocused simulations, focused beams showed larger peak to entrance ratios (26% to 38%) and focused simulations showed a two-fold increase in beam delivery efficiency. These advantages can be attributed to the magnetic acceleration of protons in the transverse plane that tends to counteract the particle outscatter that leads to degradation of peak to entrance performance in small field proton beams. Our results have important clinical implications and suggest rare earth focusing magnet assemblies are feasible and could reduce skin dose and beam number while delivering enhanced dose to narrow elongated targets (eg, in and around the spinal cord) in less time compared to collimated beams.

The effects of mapping CT images to Monte Carlo materials on GEANT4 proton simulation accuracy.

PURPOSE: Monte Carlo simulations of radiation therapy require conversion from Hounsfield units (HU) in CT images to an exact tissue composition and density. The number of discrete densities (or density bins) used in this mapping affects the simulation accuracy, execution time, and memory usage in GEANT4 and other Monte Carlo code. The relationship between the number of density bins and CT noise was examined in general for all simulations that use HU conversion to density. Additionally, the effect of this on simulation accuracy was examined for proton radiation. METHODS: Relative uncertainty from CT noise was compared with uncertainty from density binning to determine an upper limit on the number of density bins required in the presence of CT noise. Error propagation analysis was also performed on continuously slowing down approximation range calculations to determine the proton range uncertainty caused by density binning. These results were verified with Monte Carlo simulations. RESULTS: In the presence of even modest CT noise (5 HU or 0.5%) 450 density bins were found to only cause a 5% increase in the density uncertainty (i.e., 95% of density uncertainty from CT noise, 5% from binning). Larger numbers of density bins are not required as CT noise will prevent increased density accuracy; this applies across all types of Monte Carlo simulations. Examining uncertainty in proton range, only 127 density bins are required for a proton range error of <0.1 mm in most tissue and <0.5 mm in low density tissue (e.g., lung). CONCLUSIONS: By considering CT noise and actual range uncertainty, the number of required density bins can be restricted to a very modest 127 depending on the application. Reducing the number of density bins provides large memory and execution time savings in GEANT4 and other Monte Carlo packages.

In vivo iron quantification in collagenase-induced microbleeds in rat brain.

Brain microbleeds (BMB) are associated with chronic and acute cerebrovascular disease. Because BMB present in the brain is a source of potentially cytotoxic iron proportional to the volume of extravasated blood, BMB iron content is a potentially valuable biomarker both to assess tissue risk and small cerebral vessel health. We recently reported methods to quantify focal iron sources using phase images that were tested in phantoms and BMB in postmortem tissue. In this study, we applied our methods to small hemorrhagic lesions induced in the in vivo rat brain using bacterial collagenase. As expected by theory, measurements of geometric features in phase images correlated with lesion iron content measured by graphite furnace atomic absorption spectrometry. Iron content estimation following BMB in an in vivo rodent model could shed light on the role and temporal evolution of iron-mediated tissue damage and efficacy of potential treatments in cerebrovascular diseases associated with BMB.

Iron quantification of microbleeds in postmortem brain.

Brain microbleeds (BMB) are associated with chronic and acute cerebrovascular disease and present a source of pathologic iron to the brain proportional to extravasated blood. Therefore, BMB iron content is potentially a valuable biomarker. We tested noninvasive phase image methods to quantify iron content and estimate true source diameter (i.e., unobscured by the blooming effect) of BMB in postmortem human tissue. Tissue slices containing BMB were imaged using a susceptibility weighted imaging protocol at 11.7T. BMB lesions were assayed for iron content using atomic absorption spectrometry. Measurements of geometric features in phase images were related to lesion iron content and source diameter using a mathematical model. BMB diameter was estimated by image feature geometry alone without explicit relation to the magnetic susceptibility. A strong linear relationship (R(2) = 0.984, P < 0.001) predicted by theory was observed in the experimental data, presenting a tentative standardization curve where BMB iron content in similar tissues could be calculated. In addition, we report BMB iron mass measurements, as well as upper bound diameter and lower bound iron concentration estimates. Our methods potentially allows the calculation of brain iron load indices based on BMB iron content and classification of BMB by size unobscured by the blooming effect.

Quantification of punctate iron sources using magnetic resonance phase.

Iron-mediated tissue damage is present in cerebrovascular and neurodegenerative diseases and neurotrauma. Brain microbleeds are often present in these maladies and are assuming increasing clinical importance. Because brain microbleeds present a source of pathologic iron to the brain, the noninvasive quantification of this iron pool is potentially valuable. Past efforts to quantify brain iron have focused on content estimation within distributed brain regions. In addition, conventional approaches using "magnitude" images have met significant limitations. In this study, a technique is presented to quantify the iron content of punctate samples using phase images. Samples are modeled as magnetic dipoles and phase shifts due to local dipole field perturbations are mathematically related to sample iron content and radius using easily recognized geometric features in phase images. Phantoms containing samples of a chitosan-ferric oxyhydroxide composite (which serves as a mimic for hemosiderin) were scanned with a susceptibility-weighted imaging sequence at 11.7 T. Plots relating sample iron content and radius to phase image features were compared to theoretical predictions. The primary result is the validation of the technique by the excellent agreement between theory and the iron content plot. This research is a potential first step toward quantification of punctate brain iron sources such as brain microbleeds.

Correlation of hypointensities in susceptibility-weighted images to tissue histology in dementia patients with cerebral amyloid angiopathy: a postmortem MRI study.

Neuroimaging with iron-sensitive MR sequences [gradient echo T2* and susceptibility-weighted imaging (SWI)] identifies small signal voids that are suspected brain microbleeds. Though the clinical significance of these lesions remains uncertain, their distribution and prevalence correlates with cerebral amyloid angiopathy (CAA), hypertension, smoking, and cognitive deficits. Investigation of the pathologies that produce signal voids is necessary to properly interpret these imaging findings. We conducted a systematic correlation of SWI-identified hypointensities to tissue pathology in postmortem brains with Alzheimer's disease (AD) and varying degrees of CAA. Autopsied brains from eight AD patients, six of which showed advanced CAA, were imaged at 3T; foci corresponding to hypointensities were identified and studied histologically. A variety of lesions was detected; the most common lesions were acute microhemorrhage, hemosiderin residua of old hemorrhages, and small lacunes ringed by hemosiderin. In lesions where the bleeding vessel could be identified, ß-amyloid immunohistochemistry confirmed the presence of ß-amyloid in the vessel wall. Significant cellular apoptosis was noted in the perifocal region of recent bleeds along with heme oxygenase 1 activity and late complement activation. Acutely extravasated blood and hemosiderin were noted to migrate through enlarged Virchow­Robin spaces propagating an inflammatory reaction along the local microvasculature; a mechanism that may contribute to the formation of lacunar infarcts. Correlation of imaging findings to tissue pathology in our cases indicates that a variety of CAA-related pathologies produce MR-identified signal voids and further supports the use of SWI as a biomarker for this disease.

Serial susceptibility weighted MRI measures brain iron and microbleeds in dementia.

A new iron sensitive MR sequence (susceptibility weighted imaging - SWI) enabling the simultaneous quantitation of regional brain iron levels and brain microbleeds (BMB) has been acquired serially to study dementia. Cohorts of mildly cognitively impaired (MCI) elderly (n = 73) and cognitively normal participants (n = 33) have been serially evaluated for up to 50 months. SWI phase values (putative iron levels) in 14 brain regions were measured and the number of BMB were counted for each SWI study. SWI phase values showed a left putaminal mean increase of iron (decrease of phase values) over the study duration in 27 participants who progressed to dementia compared to Normals (p = 0.035) and stable MCI (p = 0.01). BMB were detected in 9 out of 26 (38%) MCI participants who progressed to dementia and are a significant risk factor for cognitive failure in MCI participants [risk ratio = 2.06 (95% confidence interval 1.37-3.12)]. SWI is useful to measure regional iron changes and presence of BMB, both of which may be important MR-based biomarkers for neurodegenerative diseases.

Fetal cerebral blood flow, electrocorticographic activity, and oxygenation: responses to acute hypoxia.

Arterial blood gases are critical in regulation of cerebral blood flow (CBF) and cerebral metabolic rate for O(2) (CMRO(2)). However, the relation of these variables to cortical tissue (t ), and electrocorticographic (ECoG) activity (high voltage low frequency, HVLF, versus low voltage high frequency, LVHF), are not well defined. In the fetus, we tested the hypothesis that ECoG pattern is associated closely with cerebral oxygenation. In fetal sheep (n = 8) with laser Doppler flowmeter, fluorescent O(2) probe and ECoG electrodes, we measured laser Doppler CBF (LD-CBF), tP(O2), ECoG and spectral edge frequency-90 (SEF(90)) in response to 40 min isocapnic hypoxia. In the normoxic fetus, LD-CBF and CMRO(2) correlated highly with ECoG state. With a shift from HVLF to LVHF, tP(O2) decreased followed by increased LD-CBF (18%) and CMRO(2) (13%). With acute hypoxia (P(aO2)= 12 +/- 1 Torr), tp(O2) decreased toapproximately 3 Torr, LD-CBF increased 48 +/- 10%, ECoG shifted to chiefly the HVLF state, SEF(90) decreased approximately 15%, and CMRO(2) decreased approximately 20% (P < 0.05 for each). For the normoxic fetus, CBF was closely related to ECoG state, but this association was less evident during acute hypoxia. We speculate that, in the otherwise stressed fetus, acute hypoxia may further compromise cerebral oxygenation.

Building Interactive Simulations in Web Pages without Programming.

A software system is described for building interactive simulations and other numerical calculations in Web pages. The system is based on a new Java-based software architecture named NumberLinX (NLX) that isolates each function required to build the simulation so that a library of reusable objects could be assembled. The NLX objects are integrated into a commercial Web design program for coding-free page construction. The model description is entered through a wizard-like utility program that also functions as a model editor. The complete system permits very rapid construction of interactive simulations without coding. A wide range of applications are possible with the system beyond interactive calculations, including remote data collection and processing and collaboration over a network.

Building interactive simulations in a Web page design program.

A new Web software architecture, NumberLinX (NLX), has been integrated into a commercial Web design program to produce a drag-and-drop environment for building interactive simulations. NLX is a library of reusable objects written in Java, including input, output, calculation, and control objects. The NLX objects were added to the palette of available objects in the Web design program to be selected and dropped on a page. Inserting an object in a Web page is accomplished by adding a template block of HTML code to the page file. HTML parameters in the block must be set to user-supplied values, so the HTML code is generated dynamically, based on user entries in a popup form. Implementing the object inspector for each object permits the user to edit object attributes in a form window. Except for model definition, the combination of the NLX architecture and the Web design program permits construction of interactive simulation pages without writing or inspecting code.