Index of content:
Volume 25, Issue 7, July 1998
Accurate segmentation and contrast measurement of microcalcifications in mammograms: A phantom study25(1998); http://dx.doi.org/10.1118/1.598302View Description Hide Description
The authors are developing a computer-aided diagnostic method to assist radiologists in differentiating between malignant and benign clustered microcalcifications in mammograms. In earlier studies we investigated shape and contrast features of microcalcifications for classification. It was found that segmentation strongly influences classification results. For this reason a phantom study has been carried out. The CDMAM phantom, consisting of a pattern of dots with known size and object contrast is used for evaluation of contrast measurement and segmentation. Dots in the range of 0.2–0.8 mm are taken as a model for microcalcifications. In this article performances of different methods for segmentation of microcalcifications are compared. An iterative method based on a Markov random field and a signal dependent criterion give satisfying results. The segmentation performances of both methods are comparable. Also the influence of the modulation transfer function on contrast estimates is determined and effect of exposure level on segmentation is analyzed.
25(1998); http://dx.doi.org/10.1118/1.598433View Description Hide Description
Monte Carlo simulations were used to quantify the amount of scattered radiation in a scanning slot detector geometry designed for use in digital mammography. Ratios of the scatter to primary (S/P) x-rayphoton energy absorbed in the detector were obtained for a Lucite phantom, and were investigated as a function of photon energy, phantom thickness, and slot detector width. Over a Lucite phantom thickness range of 2–6 cm, the S/P ratios ranged from about 0.10 to 0.17 for a 4 mm wide slot detector at the x-rayphoton energies used in mammography. These ratios increased by a factor of when the slot width was increased to 10 mm. In general, 20 keV photons gave S/P ratios similar to those of a 30 kVp x-ray spectrum (MoMo filtration). The use of a 3 cm air gap reduced the S/P ratios by a factor of between 2.5 and 3.4, depending on the phantom thickness. For a constant primary energy fluence, coherent scatter was reduced as photon energy increased, whereas Compton scatter increased with increasing photon energy. With no air gap, the contributions of coherent and Compton scatter were found to be equal at 25 keV, whereas the introduction of a 3 cm air gap resulted in equal contributions for the two scatter processes at 36 keV. A 10 mm wide slot detector consisting of a thick phosphor screen was compared to an ideal detector absorbing all incident primary/scatter photons. Average differences in the S/P ratios for these two detectors were ∼7% with no air gap and ∼4% with a 3 cm air gap. The results obtained in this study will assist in the design of an optimal slot detector for use in digital mammography.
Automatic calculation of total lung capacity from automatically traced lung boundaries in postero-anterior and lateral digital chest radiographs25(1998); http://dx.doi.org/10.1118/1.598303View Description Hide Description
Total lung capacity (TLC) is a very important parameter in the study of pulmonary function. In the pulmonary function laboratory, it is normally obtained using plethysmography or helium dilution techniques. Several authors have developed methods of calculating the TLC using postero-anterior (PA) and lateral chest radiographs. These methods have not been often used in clinical practice. In the present work, we have developed an automated computer-based method for the calculation of TLC, by determining the pulmonary contours from digital PA and lateral radiographs of the thorax. The automatic tracing of the pulmonary borders is carried out using: (1) a group of reference lines is determined in each radiograph; (2) a family of rectangular regions of interest (ROIs) defined, which include the pulmonary borders, and in each of them the pulmonary border is identified using edge enhancement and thresholding techniques; (3) removing outlaying points from the preliminary boundary set; and (4) the pulmonary border is corrected and completed by means of interpolation, extrapolation, and arc fitting. The TLC is calculated using a computerized form of the radiographic ellipses method of Barnhard. The pulmonary borders were automatically traced in a total of 65 normal radiographs (65 PA and 65 lateral views of the same patients). Three radiologists carried out a subjective evaluation of the automatic tracing of the pulmonary borders, with a finding of no error or only one minor error in 67.7% of the PA evaluations, and in 75.9% of the laterals. Comparing the automatically traced borders with borders traced manually by an expert radiologist, we obtained a precision of for the PA view, and for the lateral. The values of TLC obtained by the automatic calculation described here showed a high correlation with those obtained by applying the manual Barnhard method.
25(1998); http://dx.doi.org/10.1118/1.598304View Description Hide Description
The goal of the study was to compare a cathode-ray-tube (CRT) digital display with film by using task-dependent image quality assessment methods. Contrast-detail analysis was utilized. Human observers performed a simple detection task, specifically, detecting a pillbox target in a uniform Poisson field, using either film or a digital display that employed a CRT monitor. Observers performed equally well on both film and CRT when the window settings of the digital display were established subjectively by a radiologist. Changing the window settings of the digital display to match the average background luminance of a film-illuminator combination decreased the luminance contrast of the targets and observer performance was reduced, though these effects were probably not linked. The “gold standard” film had lower luminance contrast than the CRT displayed images, yet observer performance was never lower for film than for the CRT. Therefore we concluded that luminance contrast was not a limiting factor for observer performance in this study. The CRT monitor changed fairly rapidly after it was calibrated. During a period of six months the gamma of the display increased from 1.82 to 2.42 and the maximum luminance decreased from 319 to Low luminance output demonstrated a larger percentage decrease than high luminance output over the same time period. These observations suggest that standard window settings should be reviewed from time to time to ensure that the display is used optimally. No special look-up table setup such as perceptual linearization was used.
Optical simulations of a noninvasive technique for the diagnosis of diseased salivary glands in situ25(1998); http://dx.doi.org/10.1118/1.598305View Description Hide Description
A simulation experiment for three-dimensional (3D) imaging of exogenous fluorescinated antibodies that specifically bind to infiltrating lymphocytes in minor salivary glands was carried out. Small ( volume) rhodamine targets, which mimic diseased minor salivary glands labeled with fluorescent antibodies to infiltrating lymphocytes in Sjo/gren’s syndrome, were embedded in a highly scattering tissue phantom consisting of a thick Delrin™ disk covered by index matched Delrin™ slabs of various thickness. In this way the variation of fluorescence profiles on the surface of tissue could be examined corresponding to the range of depths of the salivary glands in vivo. Surface images were obtained for different target depths and radial distances from laser excitation to target fluorophore. These images were analyzed and compared to calculations based on random walk theory in turbid media, using previously determined scattering and absorption coefficients of the Delrin™. Excellent agreement between the surface profiles experimentally measured and those predicted by our random walk theory was obtained. Derivation of these theoretical expressions is a necessary step toward devising an inverse algorithm which may have the potential expressions to perform 3D reconstruction of the concentration distribution of fluorescent labels within tissue.
25(1998); http://dx.doi.org/10.1118/1.598306View Description Hide Description
An anthropomorphic MRS head phantom has been developed which mimics the in vivo structure, metabolite concentrations, and relaxation times (for both water and metabolites) of human braintissue. Different brain regions and two tumor types, fluid-containing ventricles, an air-filled sinus, and subcutaneous fat are all simulated. The main tissue-mimicking materials are gelatin/agar mixtures with metabolites and several other ingredients added. Their composition and method of production are thoroughly described. ’s and ’s of water in the phantom are very close to in vivo values, and metabolite ’s and ’s are considerably more realistic than those in aqueous solutions. Spectra and relaxation times for the pig brain were also acquired and compare well with those of the phantom. The realistic properties of this phantom should be useful for testing spectral quantitation and localization.
25(1998); http://dx.doi.org/10.1118/1.598307View Description Hide Description
Experiments with cultured mammallian cells exposed to ionizing radiation of varying LET are examined in the context of a microdosimetric-kinetic (MK) model of radiation induced mammalian cell killing similar to the site model of the theory of dual radiation action. The experimentally measured RBE, in the limit of zero dose, increases linearly with LET for LET less than a value that lies in the range of about 40–90 keV per micron. It is shown that the tendency of the RBE to increase linearly with LET can be explained as a result of the random variation of specific energy among microscopic sized domains (sites), into which the nucleus may be partitioned, and the effect of this variation on the formation of lethal lesions by pairwise combination of repairable primary lesions in DNA. The microscopic subunits corresponding to a site have diameter in the range of 0.56–0.75 microns. The linearity of RBE with increasing LET implies the value of the quadratic parameter of the linear-quadratic survival relation (β) is constant, at least for LET low enough to be in the linear range, and this in turn implies the production of primary DNA lesions, likely double strand breaks, does not increase with LET. The experiments, interpreted with the MK model, also imply that the repairability and potential for lethality of the primary DNA lesion does not change with variation of LET within the linear range. The failure of RBE to maintain its linear increase for higher values of LET, and the resulting RBE maximum at about 100 keV per micron, are likely primarily due to the departure of the distribution of lethal lesions among cells from the Poisson distribution. Some implications concerning the use of radiation to treatcancer are noted.
25(1998); http://dx.doi.org/10.1118/1.598308View Description Hide Description
A methodology for optimizing the beam directions in radiotherapytreatment planning has been developed and tested on a cohort of twelve prostate patients. An optimization algorithm employing an objective cost function was used, based on beam’s-eye-view volumetrics but also employing a simple dose model and biological considerations for organs-at-risk (OARs). The cost function embodies information about the volume of OARs in a single field and their position relative to the planning target volume (PTV). The proximity of the PTV to the surface of the patient is also included. Within the algorithm “importance factors” were used to model the clinical importance of different organs-at-risk so that all organs-at-risk were included in a single objective score. “Gantry-angle-windows” were introduced to restrict the available beam directions. The methodology was applied to twelve prostate patients to determine the optimum beam directions for three-field direction plans. Orientation-optimized and standard treatment plans were compared via measures of tumor control probability (TCP) and normal tissue complication probability (NTCP). Standard plans had fixed beam directions whereas orientation-optimized plans contained beam directions chosen by the algorithm. The beam-weights of both the orientation-optimized and standard plans were optimized using a dose-based simulated annealing algorithm to allow the improvements by optimizing the beam directions to be studied in isolation. The results of the comparison show that optimization of the beam directions yielded better plans, in terms of TCP and NTCP, than the standard plans. When the dose to the isocenter was scaled to produce a rectal NTCP of 1%, the average TCP of the orientation-optimized plans was greater than that for the standard plans. In conclusion, the customization of beam directions in the treatment planning of prostate patients using an objective cost function and allowed gantry-angle-windows produces superior three-field direction plans compared to standard treatment plans.
25(1998); http://dx.doi.org/10.1118/1.598296View Description Hide Description
The purpose of this work was to develop a fully automatic tool for the detection of setup deviation for small pelvic fields using, in external beam radiotherapy, an electronic portal imaging device(EPID). The algorithm processes electronic portal images of prostate cancer patients. No fiducial points or user interventions are needed. Deviation measurements are based on bone edge detection performed with the Laplacian of a Gaussian (LoG) operator. Two bone edge images are then correlated, one of which is a reference image taken as the first fraction image for the purpose of this study. The electronic portal images (EPI) also show band artefacts which are removed using the morphological top-hat transform. The algorithm was first validated with 59 phantom images acquired in clinical treatment conditions with known displacements. The algorithm was then validated with 79 clinical images where bone contours were delineated manually. For the phantom images, the setup deviations were measured with an absolute mean error of 0.59 mm and 0.47 mm with a standard deviation of 0.64 mm and 0.42 mm, horizontally and vertically, respectively. A second validation was performed using clinical prostate cancerimages. The measured patient displacements have an absolute mean error of 0.48 mm and 1.41 mm with a standard deviation of 0.58 mm and 1.30 mm in the and directions, respectively. The algorithm execution time on a SUN workstation is 5 s. This algorithm shows good potential as a setup deviation measurement tool in clinical practice. The possibility of using this algorithm combined with decision rules based on statistical observations is very promising.
25(1998); http://dx.doi.org/10.1118/1.598395View Description Hide Description
Simulation studies of a portal imaging detector based on amorphous Selenium -Se) are reported. The -Se is deposited on a metal substrate which is turned toward the megavoltage x-ray beam to act as an effective buildup layer. The thickness and material of the metal layer are studied using Monte Carlo methods by consideration of the effect of dose deposition in the -Se layer for a 6-MV beam hardened by a 20-cm polystyrene phantom. The detective quantum efficiency, due to energy absorption is modeled using the hardened primary beam. It is found that although DQE(0) increases with metal thickness up to there is a crossover near 1 cycle/mm which indicates that smaller metal thickness are more useful to visualize edges and small objects. Tungsten front plates offer the best DQE of the metals studied. The scatter fraction and scatter-to-primary ratio are also modeled to understand the effect of patient scatter on the various metal/ -Se combinations. The scatter fraction is measured experimentally for four metal plate/ -Se combinations and agrees with the Monte Carlo results within experimental uncertainties.
The indirect use of CT numbers to establish material properties needed for Monte Carlo calculation of dose distributions in patients25(1998); http://dx.doi.org/10.1118/1.598297View Description Hide Description
A number of Monte Carlo codes are available, which can be used to calculate dose distributions in patients with high accuracy. Patient geometry can readily be derived with adequate spatial resolution from CT scans. To perform the Monte Carlo calculation with the same spatial resolution, it is necessary to enter the atomic composition and density of the tissue in each voxel of the CTimage. This means entering 65 536 discrete values for a CT slice with a 256256 matrix size. The need for automated methods of setting up the material data files is obvious. Because there is no direct unique relationship between CT numbers and material composition, the aim of our work was to devise a method whereby the atomic composition and density in each voxel could be assigned automatically by indirect derivation from the CT numbers. The set of all tissues types in the human body was divided into subsets that are dosimetrically equivalent, based on Monte Carlo calculated depth dose curves in homogeneous phantoms of each tissue.CT number ranges corresponding to each tissue subset were determined from the calibration curve linking electron density with CT number for the specific CT scanner. Further subdivision was found to be necessary for the lung and bone type tissues. This was done by keeping the atomic composition constant and varying the physical density. It was found that 57 distinct tissue subsets were needed to represent the 16 main tissue types in the body at a 1% dose level. Corresponding CT number intervals of 30 HU were needed in the lung and soft tissue region, whereas in the bone region the intervals could be increased to 100 HU. A computer algorithm was set up to convert automatically from CT number to corresponding equivalent material number for the Monte Carlo preprocessor code.
An empirical relationship for determining photon beam quality in TG-21 from a ratio of percent depth doses25(1998); http://dx.doi.org/10.1118/1.598396View Description Hide Description
A key component of the Radiological Physics Center’s (RPC) on-site dosimetry review visits are photon beam calibrations for which determination of the energy of the x rays is a key element. The ratio of ionizations, for a field at depths of 20 and 10 cm for a constant SCD is used as a quantitative measure of beam quality in the Task Group 21 protocol. The RPC has measured both and the corresponding ratio of percent depth dose at a constant SSD for 685 photon beams (4–25 MV) for most makes and models of accelerators. A strong correlation between and is presented which allows the determination of the TPR ratio from the measurement of the ratio of percent depth doses. An analysis of the uncertainty introduced in the TG-21 factors ) caused by the spread in the measured data and translated into the determination of the TPR ratio results in an insignificant error This empirical relationship provides an alternate technique for quantifying the beam quality defined in the TG-21 protocol without surrendering any loss of precision in output calibration. This technique may be found by those who calibrate at a fixed SSD to be an easier and quicker method.
25(1998); http://dx.doi.org/10.1118/1.598309View Description Hide Description
Output of superficial and orthovoltage x-ray units may be measured with cylindrical or end-window parallel-plate ionization chambers. The air-kerma calibration factors for these chambers are usually determined free in air, and the x-ray machine output is stated as the air-kerma rate free in air, which, when multiplied with the appropriate backscatter factor, gives the air-kerma rate on the surface of a phantom or patient. For end-window chambers, especially when they are used for measurements of small fields or low x-ray energies, the air-kerma calibration factors may also be determined with the chamber embedded in a tissue-equivalent phantom. This results in field size dependent air-kerma in-air calibration factors but obviates the requirement for knowledge of backscatter factors when determining the air-kerma rate on the surface of a phantom. Since there still is considerable uncertainty in tabulated backscatter factors as a function of field size and x-ray beam energy, the output measurement technique which determines the air-kerma rate on phantom surface with a phantom-embedded end-window ionization chamber offers a clear advantage over the in-air calibration method.
25(1998); http://dx.doi.org/10.1118/1.598298View Description Hide Description
The PhotonRadiosurgery System is a miniature x-ray device developed for the treatment of small intracranial neoplasms. The x rays are generated at the tip of a 10-cm-long, 3-mm-diam probe with a nearly isotropic distribution. Results from measurements of the two-dimensional dose distribution around the x-ray source are presented using two methods: (1) dose measurement with an ionization chamber and a water phantom system and (2) dose measurement with radiochromic film and a solid water phantom. The shape of the two angular dose distributions in the axial plane agree with each other to within approximately 10% and the dose at 10 mm from the source, orthogonal to the probe axis, was about 20% lower than at the same distance along the axis. The relative dose difference of 20% corresponds to a change in distance from the source of at 10 mm. It is shown that the anisotropy of radiation distribution in the axial plane can be improved to approximately 10% by adjusting the electron beam with a 12% reduction in the overall radiation output.
25(1998); http://dx.doi.org/10.1118/1.598311View Description Hide Description
GAFChromic film has become increasingly popular for radiationdosimetry. In this study we explore the use of GAFChromic film as an in vivodosimeter for quality assurance (QA) of fractionated high dose rate treatments. Accuracy of dose distribution is explored for the simple vaginal cylinder geometry for which the dose can be easily calculated for comparison. Source dwell times for several patients were optimized to deliver 500 cGy at 0.5 cm from the surface of the vaginal cylinder applicator using a commercial treatment planning system. GAFChromic film was taped to the vaginal cylinder applicator and was enclosed in a leak proof rubber sleeve prior to its insertion. Optical densities were measured along the film at 2 mm spacing, using a densitometer with filtered red light. Density corrections for transient film darkening effects were made and optical densities were converted to absorbed dose in cGy. In vivo patient dose distribution measured for different patients and different fractions were compared with the calculated values along the applicator surface. The variation between the calculated and measureddose was The reproducibility of dosemeasurement for different fractions was within This study demonstrates the potential usefulness of the film as an in vivo for brachytherapy QA.
Measurements of the neutron yields from reaction (thick target) with incident energies from 1.885 to 2.0 MeV25(1998); http://dx.doi.org/10.1118/1.598299View Description Hide Description
Accelerator-basedneutron sources have been considered to be practical for boronneutron capture therapy (BNCT). Based on experience with the parameters of the Brookhaven National Laboratory BMRR reactor neutron source, which has been used in treatment experiments, the future accelerator-basedneutron source for BNCT should have the properties of low energy distribution keV) and high flux (about neutrons per second per square centimeter) in the patient zone. Using protons to bombard thick targets, generating neutrons via the reaction, is one of the optimal choices for this kind of neutron source.Neutron yield data versus incident energy are necessary in order to select the proper incident energy and for estimating how high the incident proton current should be. The required proton beam current intensity is one of the key parameters for an accelerator useful for BNCT. In the present work, neutron yields of the reaction with a thick lithium target and incident energies of 1.885 and 1.9 MeV were measured at 0° with respect to the incident beam direction. The results are and neutrons/C sr, respectively. Neutron yield angular distribution measurements at 2 MeV incident energy were also performed. The proton beams were generated by the Peking University 4.5 MV electrostatic accelerator. The emitted neutrons from these reactions have the advantages of low energy distribution and forward angular distribution, which are requirements for a BNCT neutron source. The data obtained in this work can be used as a reference to study the accelerator-basedneutron sources for BNCT.