Temporal characterization and in vitro comparison of cell survival following the delivery of 3D-conformal, intensity-modulated radiation therapy (IMRT) and volumetric modulated arc therapy (VMAT)

A phantom was designed and implemented for the delivery of treatment plans to cells in vitro. Single beam, 3D-conformal radiotherapy (3D-CRT) plans, inverse planned five-field intensity-modulated radiation therapy (IMRT), nine-field IMRT, single-arc volumetric modulated arc therapy (VMAT) and dual-arc VMAT plans were created on a CT scan of the phantom to deliver 3 Gy to the cell layer and verified using a Farmer chamber, 2D ionization chamber array and gafchromic film. Each plan was delivered to a 2D ionization chamber array to assess the temporal characteristics of the plan including delivery time and 'cell's eye view' for the central ionization chamber. The effective fraction time, defined as the percentage of the fraction time where any dose is delivered to each point examined, was also assessed across 120 ionization chambers. Each plan was delivered to human prostate cancer DU-145 cells and normal primary AGO-1522b fibroblast cells. Uniform beams were delivered to each cell line with the delivery time varying from 0.5 to 20.54 min. Effective fraction time was found to increase with a decreasing number of beams or arcs. For a uniform beam delivery, AGO-1552b cells exhibited a statistically significant trend towards increased survival with increased delivery time. This trend was not repeated when the different modulated clinical delivery methods were used. Less sensitive DU-145 cells did not exhibit a significant trend towards increased survival with increased delivery time for either the uniform or clinical deliveries. These results confirm that dose rate effects are most prevalent in more radiosensitive cells. Cell survival data generated from uniform beam deliveries over a range of dose rates and delivery times may not always be accurate in predicting response to more complex delivery techniques, such as IMRT and VMAT.

Optimization of Weighted Finite State Transducer for Speech Recognition

There is considerable interest in creating embedded, speech recognition hardware using the weighted finite state transducer (WFST) technique but there are performance and memory usage challenges. Two system optimization techniques are presented to address this; one approach improves token propagation by removing the WFST epsilon input arcs; another one-pass, adaptive pruning algorithm gives a dramatic reduction in active nodes to be computed. Results for memory and bandwidth are given for a 5,000 word vocabulary giving a better practical performance than conventional WFST; this is then exploited in an adaptive pruning algorithm that reduces the active nodes from 30,000 down to 4,000 with only a 2 percent sacrifice in speech recognition accuracy; these optimizations lead to a more simplified design with deterministic performance.

Inverse planned constant dose rate volumetric modulated arc therapy (VMAT) as an efficient alternative to five-field intensity modulated radiation therapy (IMRT) for prostate

Purpose: The aim of this work was to determine if volumetric modulated arc therapy (VMAT) plans, created for constant dose-rate (cdrVMAT) delivery are a viable alternative to step and shoot five-field intensity modulated radiation therapy (IMRT). Materials and methods: The cdrVMAT plans, inverse planned on a treatment planning system with no solution to account for couch top or rails, were created for delivery on a linear accelerator with no variable dose rate control system. A series of five-field IMRT and cdrVMAT plans were created using dual partial arcs (gantry rotating between 260° and 100°) with 4° control points for ten prostate patients with the average rectal constraint incrementally increased. Pareto fronts were compared for the planning target volume homogeneity and average rectal dose between the two techniques for each patient. Also investigated were tumour control probability and normal tissue complication probability values for each technique. The delivery parameters [monitor units (MU) and time] and delivery accuracy of the IMRT and VMAT plans were also compared. Results: Pareto fronts showed that the dual partial arc plans were superior to the five-field IMRT plans, particularly for the clinically acceptable plans where average rectal doses were less for rotational plans (p = 0·009) with no statistical difference in target homogeneity. The cdrVMAT plans had significantly more MU (p = 0·005) but the average delivery time was significantly less than the IMRT plans by 42%. All clinically acceptable cdrVMAT plans were accurate in their delivery (gamma 99·2 ± 1·1%, 3%3 mm criteria). Conclusions Accurate delivery of dual partial arc cdrVMAT avoiding the couch top and rails has been demonstrated.

Magma Evolution in the Primitive, Intra-oceanic Tonga Arc: Rapid Petrogenesis of Dacites at Fonualei Volcano

Fonualei is unusual amongst subaerial volcanoes in the Tonga arc because it has erupted dacitic vesicular lavas, tuffs and phreomagmatic deposits for the last 165 years. The total volume of dacite may approach 5 km(3) and overlies basal basaltic andesite and andesite lavas that are constrained to be less than a few millennia in age. All of the products are crystal-poor and formed from relatively low-viscosity magmas inferred to have had temperatures of 1100-1000 degrees C, 2-4 wt % H2O and oxygen fugacities 1-2 log units above the quartz-fayalite-magnetite buffer. Major and trace element data, along with Sr-Nd-Pb and U-Th-Ra isotope data, are used to assess competing models for the origin of the dacites. Positive correlations between Sc and Zr and Sr rule out evolution of the within-dacite compositional array by closed-system crystal fractionation of a single magma batch. An origin by partial melting of lower crustal amphibolites cannot reproduce these data trends or, arguably, any of the dacites either. Instead, we develop a model in which the dacites reflect mixing between two dacitic magmas, each the product of fractional crystallization of basaltic andesite magmas formed by different degrees of partial melting. Mixing was efficient because the two magmas had similar temperatures and viscosities. This is inferred to have occurred at shallow (2-6 km) depths beneath the volcano. U-Th-Ra disequilibria in the basaltic andesite and andesite indicate that the parental magmas had fluids added to their mantle source regions less than 8 kyr ago and that fractionation to the dacitic compositions took less than a few millennia. The 165 year eruption period for the dacites implies that mixing occurred on a similar timescale, possibly during ascent in conduits. The composition of the dacites renders them unsuitable candidates as contributors to average continental crust.

Recent contribution of sediments and fluids to the mantle's volatile budget

Subduction modifies the cycling of Earth's volatile elements. Fluid-rich sediments and hydrated oceanic lithosphere enter the convecting mantle at subduction zones. Some of the sediments and volatile components are released from the subducting slab, promote mantle melting and are returned to the surface by volcanism. The remainder continue into the deeper mantle. Quantification of the fate of these volatiles requires an understanding of both the nature and timing of fluid release and mantle melting(1). Here we analyse the trace element and isotopic geochemistry of fragments of upper mantle rocks that were transported to the surface by volcanic eruptions above the Batan Island subduction zone, Philippines. We find that the mantle fragments exhibit extreme disequilibrium between their U-Th-Ra isotopic ratios, which we interpret to result from the interaction of wet sediment melts and slab-derived fluids with rocks in the overlying mantle wedge. We infer that wet sediments were delivered from the slab to the mantle wedge between 8,000 and 10,000 years ago, whereas aqueous fluids were delivered separately much later. We estimate that about 625 ppm of water is retained in the wedge. A significant volume of water could therefore be delivered to the mantle transition zone at the base of the upper mantle, or even to the deeper mantle.

Rapid magmatic processes accompany arc-continent collision: the Western Bismarck arc, Papua New Guinea

New U-Th-Ra, major and trace element, and Sr-Nd-Pb isotope data are presented for young lavas from the New Britain and Western Bismarck arcs in Papua New Guinea. New Britain is an oceanic arc, whereas the latter is the site of an arc-continent collision. Building on a recent study of the Manus Basin, contrasts between the two arcs are used to evaluate the processes and timescales of magma generation accompanying arc-continent collision and possible slab detachment. All three suites share many attributes characteristic of arc lavas that can be ascribed to the addition of a regionally uniform subduction component derived from the subducting altered oceanic crust and sediment followed by dynamic melting of the modified mantle. However, the Western Bismarck arc lavas diverge from the Pb isotope mixing array formed by the New Britain and the Manus Basin lavas toward elevated Pb-208/Pb-204. We interpret this to reflect a second and subsequent addition of sediment melt at crustal depth during collision. U-238 and Ra-226 excesses are preserved in all of the lavas and are greatest in the Western Bismarck arc. High-Mg andesites with high Sr/Y ratios in the westernmost arc are attributed to recent shallow mantle flux melting at the slab edge. Data for two historical rhyolites are also presented. Although these rhyolites formed in quite different tectonic settings and display different geochemical and isotopic compositions, both formed from mafic parents within millennia.

Volumetric modulated arc therapy and breath-hold in image-guided locoregional left-sided breast irradiation

PURPOSE: To investigate the effects of using volumetric modulated arc therapy (VMAT) and/or voluntary moderate deep inspiration breath-hold (vmDIBH) in the radiation therapy (RT) of left-sided breast cancer including the regional lymph nodes.

MATERIALS AND METHODS: For 13 patients, four treatment combinations were compared; 3D-conformal RT (i.e., forward IMRT) in free-breathing 3D-CRT(FB), 3D-CRT(vmDIBH), 2 partial arcs VMAT(FB), and VMAT(vmDIBH). Prescribed dose was 42.56 Gy in 16 fractions. For 10 additional patients, 3D-CRT and VMAT in vmDIBH only were also compared.

RESULTS: Dose conformity, PTV coverage, ipsilateral and total lung doses were significantly better for VMAT plans compared to 3D-CRT. Mean heart dose (D(mean,heart)) reduction in 3D-CRT(vmDIBH) was between 0.9 and 8.6 Gy, depending on initial D(mean,heart) (in 3D-CRT(FB) plans). VMAT(vmDIBH) reduced the D(mean,heart) further when D(mean,heart) was still >3.2 Gy in 3D-CRT(vmDIBH). Mean contralateral breast dose was higher for VMAT plans (2.7 Gy) compared to 3DCRT plans (0.7 Gy).

CONCLUSIONS: VMAT and 3D-CRT(vmDIBH) significantly reduced heart dose for patients treated with locoregional RT of left-sided breast cancer. When Dmean,heart exceeded 3.2 Gy in 3D-CRT(vmDIBH) plans, VMAT(vmDIBH) resulted in a cumulative heart dose reduction. VMAT also provided better target coverage and reduced ipsilateral lung dose, at the expense of a small increase in the dose to the contralateral breast.

Accelerated partial breast irradiation (APBI): Are breath-hold and volumetric radiation therapy techniques useful?

Background: In a selective group of patients accelerated partial breast irradiation (APBI) might be applied after conservative breast surgery to reduce the amount of irradiated healthy tissue. The role of volumetric modulated arc therapy (VMAT) and voluntary moderately deep inspiration breath-hold (vmDIBH) techniques in further reducing irradiated healthy – especially heart – tissue is investigated.

Material and methods: For 37 partial breast planning target volumes (PTVs), three-dimensional conformal radiotherapy (3D-CRT) (3 – 5 coplanar or non-coplanar 6 and/or 10 MV beams) and VMAT (two partial 6 MV arcs) plans were made on CTs acquired in free-breathing (FB) and/or in vmDIBH. Dose-volume parameters for the PTV, heart, lungs, and breasts were compared. 

Results: Better dose conformity was achieved with VMAT compared to 3D-CRT (conformity index 1.24 0.09 vs. 1.49 0.20). Non-PTV ipsilateral breast receiving 50% of the prescribed dose was on average reduced by 28% in VMAT plans compared to 3D-CRT plans. Mean heart dose (MHD) reduced from 2.0 (0.1 – 5.1) Gy in 3D-CRT(FB) to 0.6 (0.1 – 1.6) Gy in VMAT(vmDIBH). VMAT is benefi cial for MHD reduction if MHD with 3D-CRT exceeds 0.5Gy. Cardiac dose reduction as a result of VMAT increases with increasing initial MHD, and adding vmDIBH reduces the cardiac dose further. Mean dose to the ipsilateral lung decreased from 3.7 (0.7 – 8.7) to 1.8 (0.5 – 4.0) Gy with VMAT(vmDIBH) compared to 3D-CRT(FB). VMAT resulted in a slight increase in the contralateral breast dose (DMean ) always remaining 1.9 Gy). 

Conclusions: For APBI patients, VMAT improves PTV dose conformity and delivers lower doses to the ipsilateral breast and lung compared to 3D-CRT. This goes at the cost of a slight but acceptable increase of the contralateral breast dose. VMAT reduces cardiac dose if MHD exceeds 0.5 Gy for 3D-CRT. Adding vmDIBH results in a further reduction of heart and ipsilateral lung dose. 

Learning Extended Tree Augmented Naive Structures

This work proposes an extended version of the well-known tree-augmented naive Bayes (TAN) classifier where the structure learning step is performed without requiring features to be connected to the class. Based on a modification of Edmonds' algorithm, our structure learning procedure explores a superset of the structures that are considered by TAN, yet achieves global optimality of the learning score function in a very efficient way (quadratic in the number of features, the same complexity as learning TANs). We enhance our procedure with a new score function that only takes into account arcs that are relevant to predict the class, as well as an optimization over the equivalent sample size during learning. These ideas may be useful for structure learning of Bayesian networks in general. A range of experiments shows that we obtain models with better prediction accuracy than naive Bayes and TAN, and comparable to the accuracy of the state-of-the-art classifier averaged one-dependence estimator (AODE). We release our implementation of ETAN so that it can be easily installed and run within Weka.