Design and implementation of a 3D-MR/CT geometric image distortion phantom/analysis system for stereotactic radiosurgery.

The design, construction and application of a multimodality, 3D magnetic resonance/computed tomography (MR/CT) image distortion phantom and analysis system for stereotactic radiosurgery (SRS) is presented. The phantom is characterized by (1) a 1 × 1 × 1 (cm)3 MRI/CT-visible 3D-Cartesian grid; (2) 2002 grid vertices that are 3D-intersections of MR-/CT-visible 'lines' in all three orthogonal planes; (3) a 3D-grid that is MR-signal positive/CT-signal negative; (4) a vertex distribution sufficiently 'dense' to characterize geometrical parameters properly, and (5) a grid/vertex resolution consistent with SRS localization accuracy. When positioned correctly, successive 3D-vertex planes along any orthogonal axis of the phantom appear as 1 × 1 (cm)2-2D grids, whereas between vertex planes, images are defined by 1 × 1 (cm)2-2D arrays of signal points. Image distortion is evaluated using a centroid algorithm that automatically identifies the center of each 3D-intersection and then calculates the deviations dx, dy, dz and dr for each vertex point; the results are presented as a color-coded 2D or 3D distribution of deviations. The phantom components and 3D-grid are machined to sub-millimeter accuracy, making the device uniquely suited to SRS applications; as such, we present it here in a form adapted for use with a Leksell stereotactic frame. Imaging reproducibility was assessed via repeated phantom imaging across ten back-to-back scans; 80%-90% of the differences in vertex deviations dx, dy, dz and dr between successive 3 T MRI scans were found to be ⩽0.05 mm for both axial and coronal acquisitions, and over >95% of the differences were observed to be ⩽0.05 mm for repeated CT scans, clearly demonstrating excellent reproducibility. Applications of the 3D-phantom/analysis system are presented, using a 32-month time-course assessment of image distortion/gradient stability and statistical control chart for 1.5 T and 3 T GE TwinSpeed MRI systems.

Adaptive Dose Escalation using Serial Four-dimensional Positron Emission Tomography/Computed Tomography Scans during Radiotherapy for Locally Advanced Non-small Cell Lung Cancer.

AIMS: Computed tomography (CT)-based radiotherapy dose escalation for locally advanced non-small cell lung cancer (LA-NSCLC) has had limited success. In this planning study, we investigated the potential for adaptive dose escalation using respiratory-gated 18F-fluorodeoxyglucose (FDG) positron emission tomography/computed tomography scans (4DPET/4DCT) acquired before and during a course of chemoradiotherapy (CRT). MATERIALS AND METHODS: We prospectively enrolled patients with LA-NSCLC receiving curative intent CRT. Radiotherapy was delivered using intensity-modulated radiotherapy (IMRT) using the week 0 4DCT scan. Three alternative, dose-escalated IMRT plans were developed offline based on the week 0, 2 and 4 4DPET/4DCT scans. The FDG-avid primary (PET-T) and nodal disease (PET-N) volumes defined by the 50% of maximum standard uptake value threshold were dose escalated to as high as possible while respecting organ at risk constraints. RESULTS: Thirty-two patients were recruited, 27 completing all scans. Twenty-five patients (93%) were boosted successfully above the clinical plan doses at week 0, 23 (85%) at week 2 and 20 (74%) at week 4. The median dose received by 95% of the planning target volume (D95) at week 0, 2 and 4 to PET-T were 74.4 Gy, 75.3 Gy and 74.1 Gy and to PET-N were 74.3 Gy, 71.0 Gy and 69.5 Gy. CONCLUSIONS: Using 18F-FDG-4DPET/4DCT, it is feasible to dose escalate both primary and nodal disease in most patients. Choosing week 0 images to plan a course with an integrated boost to PET-avid disease allows for more patients to be successfully dose escalated with the highest boost dose.

Robust PET-guided intensity-modulated radiation therapy.

PURPOSE: Functional image guided intensity-modulated radiation therapy has the potential to improve cancer treatment quality by basing treatment parameters such as heterogeneous dose distributions information derived from imaging. However, such heterogeneous dose distributions are subject to imaging uncertainty. In this paper, the authors develop a robust optimization model to design plans that are desensitized to imaging uncertainty. METHODS: Starting from the pretreatment fluorodeoxyglucose-positron emission tomography scans, the authors use the raw voxel standard uptake values (SUVs) as input into a series of intermediate functions to transform the SUV into a desired dose. The calculated desired doses were used as an input into a robust optimization model to generate beamlet intensities. For each voxel, the authors assume that the true SUV cannot be observed but instead resides in an interval centered on the nominal (i.e., observed) SUV. Then the authors evaluated the nominal and robust solutions through a simulation study. The simulation considered the effect of the true SUV being different from the nominal SUV on the quality of the treatment plan. Treatment plans were compared on the metrics of objective function value and tumor control probability (TCP). RESULTS: Computational results demonstrate the potential for improvements in tumor control probability and deviation from the desired dose distribution compared to a nonrobust model while maintaining acceptable tissue dose. CONCLUSIONS: Robust optimization can help design treatment plans that are more stable in the presence of image value uncertainties.

Poster - Thur Eve - 05: Safety systems and failure modes and effects analysis for a magnetic resonance image guided radiation therapy system.

INTRODUCTION: An online Magnetic Resonance guided Radiation Therapy (MRgRT) system is under development. The system is comprised of an MRI with the capability of travel between and into HDR brachytherapy and external beam radiation therapy vaults. The system will provide on-line MR images immediately prior to radiation therapy. The MR images will be registered to a planning image and used for image guidance. With the intention of system safety we have performed a failure modes and effects analysis. METHODS: A process tree of the facility function was developed. Using the process tree as well as an initial design of the facility as guidelines possible failure modes were identified, for each of these failure modes root causes were identified. For each possible failure the assignment of severity, detectability and occurrence scores was performed. Finally suggestions were developed to reduce the possibility of an event. RESULTS/DISCUSSION: The process tree consists of nine main inputs and each of these main inputs consisted of 5 - 10 sub inputs and tertiary inputs were also defined. The process tree ensures that the overall safety of the system has been considered. Several possible failure modes were identified and were relevant to the design, construction, commissioning and operating phases of the facility. The utility of the analysis can be seen in that it has spawned projects prior to installation and has lead to suggestions in the design of the facility.

Poster - Thur Eve - 20: Serial FDG 4DPET imaging during radiotherapy in advanced lung cancer patients.

The availability of respiratory synchronized PET (4DPET) imaging has enabled more accurate analysis of metabolic response since motion blur is minimized. We present our preliminary analysis of serial FDG 4DPET images acquired at weeks 0, 2, 4, and 7 during radiotherapy of seven stage II-III NSCLC patients. The tumor and nodal PTV of the week 0 images restrained a 4DPET image thresholding algorithm to automatically contour SUV levels ranging from 20 to 80% of the maximum SUV, creating an intensity volume histogram (IVH) for each week. These contours allowed analysis of PET volumes and standard PET metrics such as SUVmax and SUVmean . We found a trend for decreasing SUVmax and SUVmean over a treatment course in both the tumor and nodal regions. On average, the SUVmax within the tumor decreased by 17±13% (1 SD) after 2 weeks, 30±13% after 4 weeks, and 39±19% after 7 weeks of radiotherapy. Decreasing volume trends were also observed in the 20 to 80% max SUV autocontours, ranging from 26±29% to 50±40% respectively, over 7 weeks of treatment. Only one patient demonstrated an increase in FDG uptake within the tumor volume between week 0 and week 2 of treatment, and was also the only patient to recur locally at 3 months following treatment. Changes in tumor metabolism over the course of advanced NSCLC radiotherapy are quantifiable with serial FDG 4DPET imaging. Preliminary analysis suggests that variations in these trends could be useful in identifying non-responding patients that may require an alternative radiotherapeutic approach.

Poster - Thur Eve - 62: Assessing the clinical application of the van Herk margin formula for lung radiotherapy.

According to a margin recipe developed by van Herk et al. the Planning Target Volume (PTV) margin to ensure the Clinical Target Volume is covered by at least 95% of the prescribed dose can be calculated by applying the following formula: M = 2.5Σ + 1.64σ2 - 1.64σp. In the van Herk Margin formula (VHMF), Σ is the standard deviation (SD) of all systematic errors; σ is the SD of random errors and σp is the width of the penumbra. This formula is based on an idealized dose profile model that may not account for factors that vary significantly in lung radiotherapy such as tumour size and tissue density. The purpose of this study was to use accurate dose calculation algorithms and respiratory motion modeling to investigate the validity of the VHMF for lung radiotherapy. Random and systematic errors were simulated in treatment planning software using dose accumulation techniques for clinically relevant 3DCRT and IMRT treatment plans constructed on virtual phantoms. Phantom parameters such as target size, peak-to-peak motion amplitude and tissue density were varied to investigate their impact on the systematic and random error components of the margin formula. The VHMF was found to provide adequate dose coverage for all plans generated on different target sizes and motion amplitudes. Although discrepancies existed between idealized and realistic dose profiles in water and lung, the dose coverage defined by the V95 was not affected. The margin formula was found to be robust; however, further investigation of the influence of plan conformity is needed.

Early metabolic response evaluation after stereotactic radiotherapy for lung cancer: pilot experience with 18F-fluorodeoxyglucose positron emission tomography-computed tomography.

The early response of lung tumours to stereotactic radiotherapy was prospectively evaluated with 18F-fluorodeoxyglucose positron emission tomography-computed tomography. Three months after treatment, the maximum standardised uptake value and the tumour diameter fell by 64 and 30%, respectively. This imaging strategy therefore remains under ongoing evaluation with the aim of identifying predictive and prognostic factors.

Poster - Thurs Eve-22: Image guided radiation therapy for lung cancer.

OBJECTIVE: To determine the geometric accuracy of conventional and stereotactic lung radiotherapy using cone-beam CT image guidance, and assess the efficacy of these image-guided radiation therapy (IGRT) processes. MATERIALS AND METHODS: IGRT was first used for our stereotactic lung program, where high geometric accuracy is required to deliver high doses in few fractions. The initial positional accuracy for 47 patients was assessed by registering daily CBCT to the planning CT; the patient position was corrected when the CBCT indicated discrepancies > ± 3 mm in any direction. For 19 of these patients, a second CBCT was acquired to assess the residual error. IGRT was also used to assess the initial and residual errors for lung cancer patients treated conventionally with (14 pts; 584 CBCT) and without (25 pts; 1032 CBCT) a remote-controlled treatment couch. Systematic (Σ) and random (σ) positional errors were assessed for these three groups. RESULTS: For stereotactic lung patients, Σ and σ ranged between 4.1 and 6.1 mm. IGRT reduces these errors to 1.2-1.9 mm, raising the proportion of patients within ± 3 mm from 16% to 82%. For conventional lung cancer patients, Σ and σ ranged between 1.4 and 3.8 mm, and IGRT raises the proportion of patients within ± 3 mm from 27% to 67%, with the remote-controlled couch further improving this proportion to 84%. CONCLUSION: IGRT clearly confirms the high geometric accuracy required for stereotactic lung patients. This new paradigm has been transported to patients with locally-advanced lung cancer, with similar accuracy.

Trimmed radiosurgical fields.

Radiosurgery aims to deliver a high radiation dose to a small target volume while sparing surrounding healthy tissues. However, since the target volume is often large and irregularly-shaped, a significant amount of healthy tissue is irradiated. To improve conformity of the dose volume to the target volume, we propose to optimize the field shape by trimming the field described by the radiosurgery cone with the accelerator jaws for a given arc. We have measured output factors (OF), tissue-maximum ratios (TMR), off-axis ratios (OAR) and penumbrae for 40, 32.5 and 24 mm cone fields trimmed by the lower (i.e., X jaws) and /or upper (i.e., Y jaws) collimator jaws. The smallest field was 8 mm large, and length was limited by the cone size. The average penumbra due to the cone field is 2.8 mm, and 4.1 and 6.1 mm for those due to the X and Y jaws, respectively. Moreover, the penumbrae due to the X and Y jaws are independent of jaw position within the radiosurgical field. Because of the large penumbra involved with the Y jaws, radiosurgical fields should be trimmed by the X1 and/or X2 jaws only. The measured OF's have been fitted with a hyperbolic function. All of the fitted OF's fall within +/- 0.5% of the measured OF's. The TMR values obtained with trimmed fields do not change much, except for the smallest fields (up to 10% at a depth of 20 cm). Therefore, using trimmed radiosurgical fields requires straightforward dosimetric changes and provides a level of beam shaping for large cone fields (> 20 mm in diameter) without introducing additional hardware.

Optimal phosphor thickness for portal imaging.

A theoretical approach known as quantum accounting diagram (QAD) analysis has been used to calculate the spatial-frequency-dependent detective quantum efficiency (DQE) of two portal imaging systems: one based on a video camera and another based on an amorphous silicon array. The spatial frequency-dependent DQEs have then been used to determine indices of displayed and perceived image quality. These indices are figures of merit that can be used to optimize the design of linear imaging systems. We have used this approach to determine which of eight phosphor screen thicknesses (ranging between 67 and 947 mg/cm2) is optimal for the two designs of portal imaging systems. The physical characteristics (i.e., detection efficiencies, gains, and MTFs) of each of the eight x-ray detectors have been measured and combined with the physical characteristics of the remaining components to calculate the theoretical DQEs. In turn, the DQEs have been used to calculate theoretical indices of displayed and perceived image quality for two types of objects: a pelvis object and a pointlike object. The maximal indices of displayed and perceived image quality were obtained with screen thickness ranging between 358 and 947 mg/cm2, depending upon the imaging system design and the object being imaged. Importantly, the results showed that there is no single optimal screen thickness. The optimal thickness depended upon imaging task (e.g., detecting large, low-contrast structures, or detecting edges and small structures). Nevertheless, the results showed that there were only modest improvements in the indices of image quality for phosphor screens thicker than 350-400 mg/cm2.

A quantum accounting and detective quantum efficiency analysis for video-based portal imaging.

The quality of images generated with radiographic imaging systems can be degraded if an inadequate number of secondary quanta are used at any stage before production of the final image. A theoretical technique known as a "quantum accounting diagram" (QAD) analysis has been developed recently to predict the detective quantum efficiency (DQE) of an imaging system as a function of spatial frequency based on an analysis of the propagation of quanta. It is used to determine the "quantum sink" stage(s) (stages which degrade the DQE of an imaging system due to quantum noise caused by a finite number of quanta), and to suggest design improvements to maximize image quality. We have used this QAD analysis to evaluate a video-based portal imaging system to determine where changes in design will have the most benefit. The system consists of a thick phosphor layer bonded to a 1 mm thick copper plate which is viewed by a T.V. camera. The imaging system has been modeled as ten cascaded stages, including: (i) conversion of x-ray quanta to light quanta; (ii) collection of light by a lens; (iii) detection of light quanta by a T.V. camera; (iv) the various blurring processes involved with each component of the imaging system; and, (v) addition of noise from the T.V. camera. The theoretical DQE obtained with the QAD analysis is in excellent agreement with the experimental DQE determined from previously published data. It is shown that the DQE is degraded at low spatial frequencies (< 0.25 cycles/mm) by quantum sinks both in the number of detected x rays and the number of detected optical quanta. At higher spatial frequencies, the optical quantum sink becomes the limiting factor in image quality. The secondary quantum sinks can be prevented, up to a spatial frequency of 0.5 cycles/mm, by increasing the overall system gain by a factor of 9 or more, or by improving the modulation transfer function (MTF) of components in the optical chain.

Standard atlas of the gross anatomy of the developing inner ear of the chicken.

During development, the chicken inner ear undergoes a series of morphological changes which give rise to the various structures found in the adult, including the mature semicircular canals, utricle, saccule, cochlear duct, endolymphatic duct and sac, and neurons of the eighth cranial nerve ganglion. Beginning as a hollow epithelial sphere, the inner ear is sculpted into this complex labyrinth of fluid-filled ducts punctuated by their associated sensory end organs. In this report, the three-dimensional complexity of the developing inner ear of the chicken embryo is documented in the form of a standard atlas. The protocol involved fixation, dehydration, and clearing of embryonic heads harvested at daily intervals, followed by injection of an opaque dye (enamel paint suspension) into the fluid ducts of the inner ear. The position of the ear is shown relative to surface landmarks at seven different stages of development, ranging from embryonic day 5 (E5) to E18. Also shown are higher-power photomicrographs of the inner ear in isolation taken at daily intervals at E3-E17 and viewed from two orthogonal positions. Three orthogonal views are shown at 6-hour intervals during the critical stages of semicircular canal formation (E6-E7). Quantitative measurements of the linear dimensions of the inner ear (dorsoventral, anteroposterior, and mediolateral axes) as a function of time indicate a linear increase in the growth of the ear from E3 through E18. This atlas should prove valuable for evaluating mutant phenotypes in inner ear morphogenesis following gene perturbation experiments in the chicken.

Evaluation of a high-density scintillating glass for portal imaging.

One of the main factors that limits the performance of T.V. camera-based portal imaging systems is the poor light-collection efficiency of the lens and T.V. camera. An x-ray detector that produces more light per incident x ray would help overcome this limitation. We have been evaluating a high-density (3.8 g/cm3), thick (12 mm) glass scintillator for its suitability as an x-ray detector for T.V. camera-based portal imaging systems. The light output and spatial resolution of the glass scintillator has been compared to that of a copper plate/phosphor screen detector using radiographic film and the T.V. camera of our portal imaging system. The film measurements show that the light output of the glass scintillator is 82% of that of the copper plate/phosphor screen, while the T.V. camera measurements show that this value is 48%. A theoretical model of light transport described in this paper suggests that this discrepancy is due to refraction at the glass-air interface. Our measurements of the modulation transfer function (MTF) show that the spatial resolution obtained with the glass scintillator is similar to that obtained with the copper plate phosphor screen. However, the spatial resolution obtained with the glass scintillator decreases as the angle of x-ray incidence increase; this decrease, which is not observed for the copper plate/phosphor screen detector, is due to the large thickness of the glass scintillator. Due to the limited light output and the variable spatial resolution, the transparent glass scintillator, in its current form, is not suitable for portal imaging.

Limiting values of backscatter factors for low-energy x-ray beams.

Models for the calculation of upper and lower limiting values to the backscatter factor (BSF) are presented. The upper limit is obtained from Monte Carlo simulations of infinite parallel beams incident on semi-infinite phantoms with the dose contributions from all orders of photon scatter considered. The lower limits are calculated using an analytical photon transport model which considers only the primary dose and the scatter dose from photons that have undergone single scattering interactions in the phantom. The limiting values can be used to evaluate measured and modelled BSF values for x-ray beams with photons of < or = 150 keV. A parametrization of the limiting values in terms of photon energy and irradiation field size is presented so that results determined for monoenergetic beams can be extended to polyenergetic spectra. The utility of the limits is illustrated by comparisons made with BSFs from the literature.

Optimal radiographic magnification for portal imaging.

Two approaches to estimate the optimal radiographic magnification for a TV camera-based portal imaging system and portal films have been used. The first approach optimizes signal transfer while the second optimizes signal-to-noise ratio (SNR) transfer. In order to perform these optimization calculations, the physical characteristics of the imaging system (modulation transfer function and noise power spectrum) as well as the sizes of the radiation sources of our medical linear accelerators have been measured. Using these data, the optimal magnification considering signal transfer alone (M signal) has been calculated to range between 2.0 and 2.3 for the TV camera-based imaging system and is about 1.0 for portal films. Conversely, the optimal magnification considering SNR transfer (MSNR) has been calculated to range between 1.5 and 1.7 for the TV camera-based imaging system and is about 1.0 for portal films. The results suggest that most portal imaging systems are operated close to their optimal radiographic magnification.

A comparison of semiempirical models for generating tungsten target x-ray spectra.

Models for the computer generation of tungsten target x-ray spectra proposed by Birch and Marshall [Phys. Med. Biol. 24, 505-517 (1979)] and recently by Tucker, Barnes, and Chakraborty [Med. Phys. 18 211-218 (1991)] are compared. Some basic differences in the equations for the number of bremsstrahlung photons of different energies in the spectra are discussed. The models are compared in terms of their ability to characterize x-ray spectra from constant potential clinical units using three parameter equivalent spectra (EQSPEC) determined from the fit of model generated transmission curves through aluminum to measured data. The Kramers model for x-ray generation is included for completeness. It is shown that two of the models generate very similar x-ray spectra from given transmission curves although the fitting parameters in the EQSPEC characterization differ.