Hematologic toxicity can be a serious and limiting event in the treatment of squamous cell carcinoma of the anus (SCCA) with concurrent chemotherapy and radiation therapy (RT). Some groups have proposed that radiation dose to the pelvic bone marrow correlates with hematologic toxicity. We attempted to define marrow dose constraints to limit severe neutropenia and leukopenia in a multi-institutional series. Fifty-six patients with stage I-III SCCA were treated at 2 academic centers with concurrent 5-FU and mitomycin C (days 1, 28) and RT. Median age was 59 years, and median body mass index (BMI) was 24.9 kg/m2. Median RT dose was 54 Gy to the primary tumor, 45 Gy to the true pelvis, and 40 Gy to the whole pelvis. 49 patients were treated with intensity-modulated radiation therapy, and 7 patients were treated with 3D planning. The pelvic bone marrow was retrospectively contoured according to standardized guidelines and subdivided into the low pelvis, ilium, and lumbosacral spine. The V5, V10, V15, V20, V30, and V40 Gy were collected for each structure and the total marrow. Dose-volume statistics were analyzed with respect to acute grade 3 toxicity within 6 weeks of completion of treatment by univariate (UVA) and multivariable (MVA) analysis. Acute grade 3 neutropenia and leukopenia were seen in 30% and 34% of patients, respectively. On UVA, RT dose to the primary or whole pelvis was not associated with grade 3 hematologic toxicity. BMI < 25 kg/m2 was associated with grade 3 neutropenia (p < 0.01), female gender was associated with grade 3 leukopenia (p = 0.03), and mean dose to the total marrow was associated with severe leukopenia (p < 0.01) and neutropenia (p = 0.07). Exploration of various marrow parameters revealed that toxicity appeared to be driven by dose to the low pelvis more so than the ilium or lumbosacral spine. Patients who met a duad of V20 Gy ≤ 90% and V40 Gy ≤ 25% (n = 15) to the low pelvis had lower rates of grade 3 leukopenia (13% vs. 41%, p = 0.04) and neutropenia (0% vs. 40%, p < 0.01). The relationship with grade 3 leukopenia was larger for patients with a BMI < 25 kg/m2 (0% vs 55%, p < 0.01), compared to those with a BMI ≥ 25 kg/m2 (25% vs 26%, p = 0.94). The relationship with grade 3 neutropenia was also larger for patients with a BMI < 25 kg/m2 (0% vs 62%, p < 0.01), compared to those with a BMI ≥ 25 kg/m2 (0% vs 16%, p = 0.16). On MVA, the V20/40 Gy duad to the low pelvis (p < 0.01) and BMI (p < 0.01) were associated with grade 3 neutropenia, while female gender (p = 0.43) was not. All these parameters showed trends towards association with grade 3 leukopenia, including V20/40 Gy to the low pelvis (p = 0.11), BMI (p = 0.054), and female gender (p = 0.054). Acute grade 3 neutropenia and leukopenia may be limited by constraining the low pelvis bone marrow to V40 Gy ≤ 25% and a V20 Gy ≤ 90%. This guideline may be particularly important for patients with a BMI < 25 kg/m2.

In prostate cancer cases, there is no strong correlation between internal organ motion and externally visible motion, so optimal application of IGRT requires internal tracking techniques. Real-time 3D prostate tracking based on combined MV-kV imaging is an attractive solution since it uses the built-in hardware of modern LINACs equipped with OBI and EPID. However, continuous kV imaging throughout the treatment can lead to excess diagnostic dose exposure to the patient. To manage kV exposure, this work investigates the incorporation of the delivered kV tracking dose at the treatment planning stage through combined MV-kV beam modeling and conformal kV beam aperture shaping. A Monte Carlo scheme was developed to model dose deposition kernels for a 125 kVp photon imaging source, which were imported into a treatment planning system. The kV imaging source, orthogonal to the MV source from a LINAC, was commissioned in the TPS, using dose profile and PDD measurements made on radiographic film and calibrated with ion chamber measurements. Using previously treated 9-field IMRT prostate patient treatment plans, the TPS was used to calculate the dose that would be delivered by the kV beam during real-time tracking. To provide stereoscopic imaging at each gantry angle during treatment, the kV and MV beam times were matched. Apertures studied included a small 3 x 3 cm field capable of tracking closely placed fiducials in prostate cases, as well as a number of larger apertures. In the four patients studied, the average dose to the prostate from a conformed 3x3 cm kV aperture was 5 cGy. This increased to 9 cGy per fraction when using an open aperture. For the patients studied, this corresponds to 2-5% of the prescribed dose per fraction. The 3x3 cm aperture significantly spared more healthy tissue, delivering 5 cGy or more per fraction to less than 1% of a patient volume, compared to 80% of the patient volume in the full kV field case. The use of conformal kV apertures shaped to the tracking region of interest can lead to significant reductions in total kV diagnostic radiation delivered. This allows for the kV beam to provide useful tracking information, while simultaneously delivering therapeutic dose to the prostate.

Post mastectomy chest wall patients receive radiotherapy using 3D conformal or intensity modulated radiotherapy (IMRT) photon beams. The treatments can be further improved by adding modulated electron radiotherapy (MERT) components to spare the distal organs at risks (OAR). This study reports the implementation of a technique to combine MERT and IMRT fields within Eclipse treatment planning system (TPS). A post mastectomy patient previously planned and treated using photon IMRT for the right chest wall including bilateral internal mammary lymph nodes was re-planned using combined electron-photon IMRT (epIMRT). For the MERT component, distances from the external contour to the distal end of the PTV were calculated and projected on a beams-eye view. Electron segment energies were selected using the 80% isodose line (cm) = E(MeV)/3 rule, and an electron energy map was generated. The electron segments were then shaped using this energy map. The resulting 6 electron block segments for 5 energies were then imported into Eclipse. The electron dose was calculated using the electron Monte Carlo (eMC) dose calculation module. The output of each segment was adjusted to minimize the OAR dose while increasing the PTV coverage. The MERT plan was then used as a base dose in photon IMRT planning in Eclipse. The dose for photon IMRT was calculated and added to the MERT dose and compared against the photon alone IMRT plan. The epIMRT combination plan had 6 electron and 8 photon IMRT fields, contributing 40% and 60% of the mean PTV dose respectively. The epIMRT resulted in the same PTV coverage as the photon IMRT alone plan while reducing OAR doses. The mean heart dose was reduced by 23% (from 14.1 Gy to 10.9 Gy) with epIMRT. The mean ipsilateral lung dose was reduced from 23.2 Gy to 22.6 Gy, whereas mean contralateral lung dose was reduced from 5.5 Gy to 4.6 Gy with epIMRT. Electron-photon combined IMRT (epIMRT) planning was successfully implemented in Eclipse TPS. A combination of electron fields that are shaped with blocks and photon IMRT fields provided the same PTV coverage while reducing the OAR dose, especially for the heart. Although this approach increases the complexity of the plan, some patients such as post mastectomy chest wall patients, can benefit from this type of combination treatment. An intensity modulated electron and photon therapy may supplement the photon IMRT programs especially for treatment of shallow tumors.

Introduction: Vascular permeability and perfusion maps using dynamic-contrast-enhanced MRI (DCE-MRI) were performed to characterize the response of a radioresistant cell line to therapy. Ionizing radiation (IR) is a staple for the treatment of malignant tumors. However, failure to cure tumors is thought to be due to an intrinsic tumor cell radioresistance and the tumor’s microenvironment. Nu61 is a radioresistant tumor derived from a human head and neck squamous cell carcinoma (SCC-61) parental line. A genetically modified adenoviral vector (Ad.EGR-TNFα) which causes infected cells to produce tumor necrosis factor alpha (TNFα), is upregulated by a radiation inducible promoter, only when those cells have been irradiated. TNFα is a potent antitumor and antivascular agent, with a variety of potential undesirable side effects on normal tissues. By limiting its expression only to the areas irradiated, the systemic side effects may be minimized, while the synergistic effect of irradiation and TNFα is exploited. In this pilot study, different treatments were evaluated and characterized with DCE-MRI. Methods: Female athymic nude mice with Nu61 or SCC-61 xenografts in the right hind limb were randomized into one of the following groups: I) Ad.null (Nu61=6, SCC-61=3), II) Ad.null + 10 Gy (Nu61=3, SCC-61=4), III) Ad.Egr-TNF (Nu61=3, SCC-61=4) or IV) Ad.Egr-TNF + 10 Gy (Nu61=4, SCC-61=4), where Ad.null is a control vector. On the initial treatment day, MRI was performed, followed by injection of either vector plus 10 Gy of irradiation, 2-3 hours post injection (depending on the group). Three days later, the mice were again imaged. No additional treatment was given after day 0. All mice were anesthetized during all procedures. Images were acquired at 9.4 T, using a Bruker MRI Scanner with a custom 10-leg low-pass, volume birdcage coil around the tumor bearing leg. For anatomical guidance, the multislice, Rapid Acquisition with Relaxation Enhancement (RARE) spin-echo sequence (TR = 4000 ms, effective TE = 28 ms, FOV = 2.56 cm, matrix size = 256 × 256, slice thickness = 0.75 mm, NEX = 2, RARE factor = 4) was used. For DCE-MRI, T1 weighted images were acquired using a Fast Low Angle Shot (FLASH) gradient-echo sequence (TR/TE = 40 ms/3.6 ms, flip angle = 20 ̊, FOV = 2.75 cm, matrix size = 128 × 128, slice thickness = 1.5 mm, NEX = 1), with temporal resolution of 5 sec. Contrast agent (CA) (OmniScan) kinetics were fit to a two-compartment model, where K is a measure of the rate of uptake of CA and ve is the volume accessible to the CA. An ANOVA was performed to determine significance and a level of p < 0.05 was considered significant. All data are presented as mean ± SEM. Results: K and ve (in all regions of interest) were not significantly different among the four treatment groups with Nu61 (p>0.5). The large difference in K with Nu61 cell line is likely due to an outlier in the Ad.Egr-TNF + 10 Gy treated group. However, the tumor region of the mice with the SCC-61 xenografts had a significantly lower K 3 days after Ad.Egr-TNF + 10 Gy, when compared to Ad.null or Ad.null + 10 Gy (45% and 50%, respectively with p=0.03). In the tumor region, ve was 64% lower 3 days after treatment when comparing Ad.null alone with Ad.Egr-TNF + 10 Gy (p=0.006). The results are summarized in Table 1 for the tumor region, 3 days after therapy. Conclusion: Kinetic modeling of DCEMRI data suggest that Nu61 tumors, which are radioresistant, are also resistant to TNFα therapy. However the parent cell line, SCC61 responds to TNFα and irradiation via an apparent decrease in permeability and perfusion. More experiments are in progress to support these promising preliminary results, e.g. positron emission tomography is also being used to characterize these two cell lines. The radiation-induced gene therapy used here is currently undergoing clinical trials. DCE-MRI could be used to guide subsequent fractions in an adaptive image guided approach, possibly identifying non-responding regions. Table 1: DCE-MRI Results, 3 Days Post Treatment K [min] ve

Increased frequency of CT imaging can increase setup accuracy and prevent misplaced therapeutic dose due to interfraction organ motion. This is especially important for sites in which the motion can be large and is unpredictable, such as the prostate. Adaptive radiation therapy (ART) with daily imaging is one method to increase the accuracy of dose delivery to the prostate. However with increased CT frequency there is also an increase in non-targeted dose from the kV x-rays used for imaging. A promising dose reduction strategy is conformal region of interest (ROI) imaging, whereby only the target and immediately surrounding structures are illuminated from any projection angle. This leads to an interior reconstruction problem for which there is currently no stable reconstruction technique but is an active area of research. Alternatively, instead of blocking the rays that do not directly illuminate the ROI they can be heavily filtered at the source, thereby keeping patient exposure low but still providing information allowing for stable reconstruction. This intensity-weighted method also provides sufficient anatomic data in the outer region for image registration, which is critical for accurate dose summation in an ART treatment scheme. Intensity-weighted ROI imaging varies the spatial beam intensity by introducing filters in the kV x-ray beam. The results reported here are for 3mm copper filters which the authors have previously shown allows for full volume reconstruction with sufficient SNR to enable accurate rigid body registration of the skeletal structure. For conformal imaging radiopaque lead blades were used to shape the beam. The imaged ROI was taken to be a 10cm diameter cylinder along the central axis, which for an average patient is expected, on any given day, to include the entire prostate and the highest dose regions of the bladder and rectum. Two-dimensional dose distributions were measured using Gafchromic film placed between slabs 33 and 34 of a standard anthropomorphic RANDO Man phantom (The Phantom Laboratory, Salem, NY). Conformal ROI imaging was able to reduce the dose to the region outside the ROI by 50-70% and the dose to the ROI by 25% due to reduced scatter from the outer region. Intensity-weighted ROI imaging achieved 75-90% of the dose reduction of conformal imaging in the outer region while reducing dose to the inner region only half to two-thirds as much. Measurements indicate that for prostate imaging intensity modulation can provide nearly as much dose reduction as conformal illumination while still providing information of the surrounding structure, enabling stable reconstruction and registration required for accurate dose summation for ART treatment schemes.

Regions of tumors with low pO2 are important determinants of their biology and response to therapy. Electron paramagnetic resonance oxygen images (EPROI) provide accurate spatially resolved pO2 distributions in tumors and normal tissue. We explored the relationship between EPROI pO2 distributions and registered localized tumor biopsy content of vascular endothelial growth factor (VEGF) in mouse fibrosarcomas. The EPROI of the legs of C3H mice bearing FSa fibrosarcomas were obtained using a well published continuous wave (CW) technique in 48 minutes. Localized biopsies were obtained stereotactically registered with the EPROI using a #11 bone biopsy needle. This volume contained approximately 100 image voxels. The VEGF concentrations assayed using ELISA were compared with mean pO2, median pO2, and the fraction of voxels in the biopsy volume with pO2 less than 3, 6, and 10 torr (HF10). All pO2 variables correlated with VEGF concentrations (p = 0.0049 to 0.019). The most significant correlation was with the fraction of voxels in the biopsy volume with pO2 less than 10 torr. This further validates EPROI hypoxic fractions at the molecular level. The EPROI obtained in this way measures chronic hypoxia. The correlation with VEGF protein concentration is most likely with chronic hypoxia. The EPROI provide a new paradigm for the assessment of the response of tumor cells to their environment and a new direction for the evaluation of hypoxia in human tumors and planning cancer therapies.

To review outcomes of salvage radiation therapy (RT) after prostatectomy, comparing Intensity-Modulated RT (IMRT) with conformal/conventional (non-IMRT) techniques. Eighty-three patients with localized prostate cancer and a rising PSA after prostatectomy were treated at the University of Chicago between 1988 and 2005. Median pre-RT PSA was 0.64 ng/mL, 72% had pT3 disease, 12% had Gleason 8 disease or higher, and surgical margins were positive in 56%. Median RT dose was 66 Gy (range 46–76); 34 patients were treated with IMRT. 27 patients also had androgen deprivation therapy for a median 4 months. Biochemical failure was defined as any PSA >0.2 ng/mL with any consecutive rise thereafter. Median follow-up was 43 months, and was shorter for IMRT patients (median 29 months) than for non-IMRT patients (median 64 months, p < 0.0001). Toxicity was scored using RTOG criteria for gastrointestinal (GI) and genitourinary (GU) systems. At 4-years, freedom from biochemical failure (FFbF) was 53%. Margin status was the only variable associated with FFbF on univariate (p = 0.0157) and multivariate analysis (p = 0.0231) including pT3, Gleason score ≥4 + 3, hormone use, pre-RT PSA, and IMRT. Because whole pelvic treatment was associated with a higher rate of acute (p = 0.0555) or late Grade 2+ toxicity (56% vs 16%, p = 0.0124), 9 such patients were excluded for comparison between IMRT and non-IMRT (Table 1). Dose-volume histogram (DVH) data was available for 28 of 73 patients. No tested DVH parameters were associated with acute or late grade 2+ GI or GU toxicity. A comparison of DVH data for rectal tissue demonstrated that IMRT increased hot spots [V70 7% vs 0%, p = 0.0050, dose to hottest 10% of rectum 69 vs 67 Gy, p = 0.0117], reduced the volume of rectal tissue receiving medium doses of RT [V60, V40 lower for IMRT, p > 0.05], and increased the volume receiving lower doses [V20, V10 higher for IMRT, p > 0.05]. For bladder tissue, IMRT increased hot spots [V70 70% vs 67%, p = 0.0005, dose to hottest 10% of tissue 70 vs 65 Gy, p = 0.0004] but did not better spare bladder from medium or lower doses of RT [equal or higher V60, V40, V20, V10, p > 0.05]. Salvage RT is effective in select patients after prostatectomy. Even for limited prostate bed volumes, IMRT may reduce rectal tissue receiving 30–60 Gy and improve acute GI toxicity; however, this comes at the expense of increased hot spots to the rectum and bladder which may impact late toxicity. Further study is necessary to better assess the risk-benefit ratio of IMRT in this setting.Table 1Toxicity following salvage RT to the prostatic bedOverall (n = 73)IMRT (n = 31)Non-IMRT (n = 42)p-valueAcute ≥ Grade 2 GI20/72 (28%)13%38%0.0170 ≥ Grade 2 GU5/73 (7%)3%10%0.2719Late ≥ Grade 2 GI5/69 (7%)10%5%0.4828 ≥ Grade 2 GU11/69 (16%)26%8%0.0417 ≥ Grade 2 GI or GU15/69 (22%)32%13%0.0552 ≥ Grade 3 GI or GU3/69 (4%)10%0%0.0258 Open table in a new tab

The use of spin echoes to obtain spectroscopic EPR images (spectral–spatial images) at 250 MHz is described. The advantages of spin echoes—larger signals than the free induction decay, better phase characteristics for Fourier transformation, and decay shapes undistorted by instrumental dead time—are clearly shown. An advantage is gained from using a crossed loop resonator that isolates the 250‐W pump power by greater than 50 dB from the observer arm preamplifiers. The echo decay rates can be used to determine the oxygen content in solutions containing 1 mM trityl concentrations. Two‐ and three‐dimensional images of oxygen concentration are presented. Magn Reson Med, 2006. © 2006 Wiley‐Liss, Inc.

Purpose: To develop the helical cone‐beam scanning capability on a simulator CT system and apply a recently developed backprojection‐filtration (BPF) algorithm to reconstruct exact 3D images from data obtained in this system. Method and Materials: Helical cone‐beam CT is an emerging technology with many advantages over conventional CT scans, such as fast volume coverage speed, efficient use of x‐ray power, and sufficient data for exact 3D image reconstruction. We developed the helical cone‐beam capability on a simulator CT device (Acuity, Varian Medical Systems), which includes a kV x‐ray source, a patient couch, and an amorphous silicon flat‐panel detector (Varian PaxScan 4030CB). We fabricated a small motorized leadscrew mechanism on the couch of the Acuity, which allows for longitudinal translation of the patient couch at a constant speed as the gantry rotates. We applied the BPF algorithm to reconstruct exact 3D images from data obtained in this system. Results: A CatPhan phantom was used for data acquisition. 683 projection data were collected during a one‐turn gantry rotation while the couch was translated longitudinally with a helical pitch of 187.2 mm. The x‐ray source was operated at 125 kVp and 560 mAs. Before projection data were used for reconstruction, a number of corrections were performed, such as bad pixel, dark field, flood field, detector sag. We also applied simple corrections for beam hardening and scattering. 3D images were reconstructed from the corrected data by use of the BPF algorithm. Conclusion: We have developed for the first time the helical scanning capability on a simulator cone‐beam CT system and performed a preliminary phantom study. The BPF algorithm was used to reconstruct exact 3D images from the helical cone‐beam data. Such approach can readily be extended to cone‐beam CT components on a LINAC to provide more accurate image representations of the patient.

A method has been developed which allows accurate registration of 3D image data sets of the head, such as CT or MRI, with with the anatomy of the actual patient. Once registration is accomplished, the patient and image spaces may be interactively explored, and any point or volume of interest in either space instantly transformed to the other. This paper demonstrates the use of this technology in accurately transferring radiation therapy treatment plans from the 3D image space in which they are simulated, to the physical patient. This provides a heretofore missing objective link between 3D image-based simulations and actual treatment delivery.

It is well known that cone‐beam data acquired with a circular orbit are insufficient for exact image reconstruction. Despite this, because a cone‐beam scanning configuration with a circular orbit is easy to implement in practice, it has been widely employed for data acquisition in, e.g., micro‐CT and CT imaging in radiation therapy. The algorithm developed by Feldkamp, Davis, and Kress (FDK) and its modifications, such as the Tent–FDK (T‐FDK) algorithm, have been used for image reconstruction from circular cone‐beam data. In this work, we present an algorithm with spatially shift‐variant filtration for image reconstruction in circular cone‐beam CT. We performed computer‐simulation studies to compare the proposed and existing algorithms. Numerical results in these studies demonstrated that the proposed algorithm has resolution properties comparable to, and noise properties better than, the FDK algorithm. As compared to the T‐FDK algorithm, our proposed algorithm reconstructs images with an improved in‐plane spatial resolution. © 2005 Wiley Periodicals, Inc. Int J Imaging Syst Technol, 14, 213–221, 2004; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/ima.20026

In radiation therapy for breast cancer treatment, information about the external (skin) and internal (lung) boundaries is highly useful for determining the relative locations of the target and lung. In this work., we investigate the feasibility of tomographic reconstruction from few-view and limited-angle cone-beam projections acquired in radiation therapy unit for obtaining critical boundary information. From the few-view and limited-angle projections acquired directly in the treatment machine with an amorphous silicon electronic portal imaging device (EPID), We compared and evaluated the performance of the conventional cone-beam FDK algorithm and an iterative algorithm based upon the maximum-likelihood method for transmission tomography (ML-TR). Preliminary results demonstrated that the ML-TR algorithm is more promising than is the cone-beam FDK algorithm. Useful boundary information for breast localization can be obtained with very few projections in a limited angle range from the reconstruction of ML-TR algorithm.

We present an algorithm for image reconstruction from data acquired by use of an asymmetric fan-beam configuration with a spatially varying focal length. This algorithm is a natural generalization of our previous algorithm for fan-beam computed tomography (CT) that involves shift-variant filtering, and it possesses desirable noise and numerical properties similar to those in the previous algorithm. We performed computer simulation studies to validate and evaluate the proposed algorithm, and quantitative results demonstrate that the proposed algorithm possesses desirable noise and numerical properties. The filtered backprojection algorithm can reconstruct images from data acquired in a configuration with a constant focal length. In this situation, the proposed algorithm yields images with lower and more uniform noise properties than the filtered backprojection algorithm. (C) 2004 Society of Photo-Optical Instrumentation Engineers.

We propose a generalized algorithm for image reconstruction from data acquired in an asymmetric fan-beam configuration with a spatially varying focal length. This algorithm is an extension of our previous algorithm for fan-beam computed tomography (CT) that involves a shift-variant filtering. As compared to the existing algorithms, the proposed algorithm involves no additional explicit interpolation and eliminates the spatially-variant weighting factor. Therefore, it retains the spatial resolution and is less susceptible to data noise. We performed computer-simulation studies to validate and evaluate the proposed algorithm, and quantitative results demonstrate that the proposed algorithm possesses desirable noise and numerical properties.

This work presents a methodology for obtaining quantitative oxygen concentration images in the tumor-bearing legs of living C3H mice. The method uses high-resolution electron paramagnetic resonance imaging (EPRI). Enabling aspects of the meth odology include the use of injectable, narrow, single-line triaryl methyl spin probes and an accurate model of overmodulated spectra. Both of these increase the signal-to-noise ratio (SNR) resulting in high resolution in space (1 mm)(3) and oxygen con centrations (similar to3 torr). Thresholding at 15% the maximum spectral amplitude gives leg/tumor shapes that reproduce those in photographs. The EPRI appears to give reasonable oxygen partial pressures, showing hypoxia (similar to0-6 torr, 0-10(3) pa) in many of the tumor voxels. EPRI was able to detect statistically significant cant changes in oxygen concentrations in the tumor with ad ministration of carbogen, although the changes were not in creased uniformly. As a demonstration of the method, EPRI was compared with nearly concurrent (same anesthesia) T*(2)/blood oxygen level-dependent (BOLD) MRI. There was a good spatia correlation between EPRI and MRI. Homogeneous and heterogeneous T*(2)/BOLD MRI correlated well with the quantitative EPRI. This work demonstrates the potential for EPRI to display at high spatial resolution, quantitative oxygen tension changes in the physiologic response to environmental changes. (C) 2003 Wiley-Liss, Inc.

Introduction: One of the major factors, which limit our ability to take full advantage of closely tailored field arrangements in the delivery of external beam radiation for prostate cancer, is the dilemma of targeting errors due to both daily internal organ motion and setup errors between the time of simulation and treatment. The current randomized phase III trial is designed to test the hypothesis that tighter margins achievable through daily 3D ultrasound target localization will result in an improvement in treatment related complications.

We developed a novel algorithm for image reconstruction from fan-beam data acquired with asymmetric flat-panel detectors. This new algorithm can improve the noise properties of the widely used fan-beam filtered-backprojection (FFBP) algorithm by eliminating the spatially-variant weighting factor while retaining FFBP's favorable resolution properties. Quantitative results verify the theoretical prediction of the. improved, noise properties. These improved noise properties can be translated into a reduction of radiation dose. The new algorithm is particularly robust and useful when applied to CT systems with large field of measurement (FOM) and/or relatively small focal lengths.

The accuracy of a surface fitting technique was investigated for registration of SPECT (single photon emission computed tomography) brain images. Fiduciary markers were placed on the outside of a 3-D brain phantom filled with Tc-99m solution. The phantom and markers were scanned at several orientations in the scanner, and contours were produced for the slices of each set. Parameters were fit to rotate and translate images acquired at each orientation to match the images acquired at each of the other orientations. Independent fits based on the measured marker positions were done to determine the rotational and translational parameters more precisely. The maximum error measured in the translational parameters was less than 1/2 pixel (1.8 mm) in any direction and the errors in the angle parameters were all less than 3 degrees, with most less than 1.5 degrees.<>