Showing papers in "Medical Physics in 1996"
TL;DR: The quantitative use of susceptibility data is important to MRI, but the use of literature values for the susceptibility of materials is often difficult because of inconsistent traditions in the definitions and units used for magnetic parameters-particularly susceptibility.
Abstract: The concept of magnetic susceptibility is central to many current research and development activities in magnetic resonance imaging (MRI); for example, the development of MR-guided surgery has created a need for surgical instruments and other devices with susceptibility tailored to the MR environment; susceptibility effects can lead to position errors of up to several millimeters in MR-guided stereotactic surgery; and the variation of magnetic susceptibility on a microscopic scale within tissues contributes to MR contrast and is the basis of functional MRI. The magnetic aspects of MR compatibility are discussed in terms of two levels of acceptability: Materials with the first kind of magnetic field compatibility are such that magnetic forces and torques do not interfere significantly when the materials are used within the magnetic field of the scanner; materials with the second kind of magnetic field compatibility meet the more demanding requirement that they produce only negligible artifacts within the MR image and their effect on the positional accuracy of features within the image is negligible or can readily be corrected. Several materials exhibiting magnetic field compatibility of the second kind have been studied and a group of materials that produce essentially no image distortion, even when located directly within the imaging field of view, is identified. Because of demagnetizing effects, the shape and orientation, as well as the susceptibility, of objects within and adjacent to the imaging region is important in MRI. The quantitative use of susceptibility data is important to MRI, but the use of literature values for the susceptibility of materials is often difficult because of inconsistent traditions in the definitions and units used for magnetic parameters-particularly susceptibility. The uniform use of SI units for magnetic susceptibility and related quantities would help to achieve consistency and avoid confusion in MRI.
1,408 citations
TL;DR: A new model for calculating electron beam dose based on a two- or three-dimensional geometry defined by computerized tomography images, based on the Voxel Monte Carlo model (VMC), was tested in comparison to calculations by EGS4 and the "Hogstrom algorithm" (MDAH) using several fictive phantoms.
Abstract: A new model for calculating electron beam dose has been developed The algorithm is based on a two- or three-dimensional geometry defined by computerized tomography (CT) images The Monte Carlo technique was used to solve the electron transport equation However, in contrast to conventional Monte Carlo models (EGS4) several approximations and simplifications in the description of elementary electron processes were introduced reducing in this manner the computational time by a factor of about 35 without significant loss in accuracy The Monte Carlo computer program does not need any precalculated data The random access memory required is about 16 Mbytes for a 128(2) X 50 matrix, depending on the resolution of the CT cube The Voxel Monte Carlo model (VMC) was tested in comparison to calculations by EGS4 and the "Hogstrom algorithm" (MDAH) using several fictive phantoms In all cases a good coincidence has been found between EGS4 and VMC, especially near tissue inhomogeneities, whereas the MDAH algorithm has produced dose underestimations of up to 40%
323 citations
TL;DR: Tests of that model's "stationarity" assumptions are illustrated using radiologists' free-search interpretations of chest films for pulmonary nodules, and they suggest that human observers may often violate those assumptions when making multiple-report interpretations of images.
Abstract: In this paper methods used to measure observer performance are reviewed, and a simple general model for finding and reporting target objects in gray-scale image backgrounds is presented. That model provides the basis for a combined measurement of detection and localization performance in various image-interpretation tasks, whether by human observers or by realized computer algorithms. The model assumes that (1) an observer's detection response and first choice of target location both depend on the "maximally suspicious" finding on an image, (2) a correct (first-choice) localization of the actual target occurs if and only if its location is selected as the most suspicious, and (3) a target's presence does not alter the degree of suspicion engendered by any other (normal) image findings. Formalization of these assumptions relates the ROC curve, which measures the ability to discriminate between images containing targets and images without targets, to the "Localization Response" (LROC) curve, which measures the conjoint ability to detect and correctly localize the actual targets in those images. A maximum-likelihood statistical procedure, developed for a two-parameter "binormal" version of this model, concurrently fits both the ROC and LROC curves from an observer's image ratings and target localizations for a set of image interpretations. The model's application is illustrated (and compared to standard ROC analysis) using sets of rating and localization data from radiologists asked to search chest films for pulmonary nodules. This model is then extended to multiple-report ("free-response") interpretations of multiple-target images, under the stringent requirement that an observer's detection capability and criterion for reporting possible targets both remain stationary across images and across the successive reports made on a given image. That extended model yields formulations and predictions for the so-called "Free-Response" (FROC) curve, and for a recently proposed "Alternative FROC" (AFROC) curve. Tests of that model's "stationarity" assumptions are illustrated using radiologists' free-search interpretations of chest films for pulmonary nodules, and they suggest that human observers may often violate those assumptions when making multiple-report interpretations of images.
305 citations
TL;DR: There is excellent agreement between the dose distributions predicted using treatment planning calculations and those measured using the gel method, and the clinical practical utility of MRI-based polymer gel dosimetry is thereby demonstrated.
Abstract: Further progress in the development of polymergeldosimetry using MRI is reported, together with examples of its application to verify treatment plans for stereotactic radiosurgery and high dose rate brachytherapy. The dose distribution image produced in the tissue‐equivalent gel by radiation‐induced polymerization, and encoded in the spatial distribution of the NMR transverse relaxation rates (R 2) of the water protons in the gel, is permanent. Maps of R 2 are constructed from magnetic resonance imaging data and serve as a template for dose maps, which can be used to verify complex dose distributions from external sources or brachytherapy applicators. The integrating, three‐dimensional, tissue‐equivalent characteristics of polymergels make it possible to obtain dose distributions not readily measured by conventional methods. An improved gel formulation (BANG‐2) has a linear dose response that is independent of energy and dose rate for the situations studied to date. There is excellent agreement between the dose distributions predicted using treatment planning calculations and those measured using the gel method, and the clinical practical utility of MRI‐based polymergeldosimetry is thereby demonstrated.
291 citations
TL;DR: It is demonstrated that accurate and rapid delivery of realistic intensity-modulated plans is feasible using a dynamic multi-leaf collimator and the impact of the "tongue-and-groove" effect on dynamic MLC treatments is studied and it is possible to render this effect inconsequential by appropriately synchronizing leaf motions.
Abstract: The optimization of intensity distributions and the delivery of intensity‐modulated treatments with dynamic multi‐leaf collimators(MLC) offer important improvements to three‐dimensional conformal radiotherapy. In this study, a nine‐beam intensity‐modulated prostate plan was generated using the inverse radiotherapy technique. The resulting fluence profiles were converted into dynamic MLC leaf motions as functions of monitor units. The leaf motion pattern data were then transferred to the MLCcontrol computer and were used to guide the motions of the leaves during irradiation. To verify that the dose distribution predicted by the optimization and planning systems was actually delivered, a homogeneous polystyrene phantom was irradiated with each of the nine intensity‐modulated beams incident normally on the phantom. For each exposure, a radiographicfilm was placed normal to the beam in the phantom to record the deposited dose. The films were calibrated and scanned to generate 2‐D isodose distributions. The dose was also calculated by convolving the incident fluence pattern with pencil beams. The measured and calculated dose distributions were compared and found to have discrepancies in excess of 5% of the central axis dose. The source of discrepancies was suspected to be the rounded edges of the leaves and the scattered radiation from the various components of the collimation system. After approximate corrections were made for these effects, the agreement between the two dose distributions was within 2%. We also studied the impact of the ‘‘tongue‐and‐groove’’ effect on dynamic MLCtreatments and showed that it is possible to render this effect inconsequential by appropriately synchronizing leaf motions. This study also demonstrated that accurate and rapid delivery of realistic intensity‐modulated plans is feasible using a dynamic multi‐leaf collimator.
191 citations
TL;DR: Data files from Lawrence Livermore National Laboratory were truncated to match the needs of the medical physics community, and an interpolation routine was written to calculate a continuous set of cross sections spanning energies from 1 keV to 50 MeV.
Abstract: The purpose of this technical report is to make available an up-to-date source of attenuation coefficient data to the medical physics community, and to compare these data with other more familiar sources. Data files from Lawrence Livermore National Laboratory (in Livermore, CA) were truncated to match the needs of the medical physics community, and an interpolation routine was written to calculate a continuous set of cross sections spanning energies from 1 keV to 50 MeV. Coefficient data are available for elements Z = 1 through Z = 100. Values for mass attenuation coefficients, mass-energy-transfer coefficients, and mass-energy absorption coefficients are produced by a single computer subroutine. In addition to total interaction cross sections, the cross sections for photoelectric, Rayleigh, Compton, pair, and some triplet interactions are also produced by this single program. The coefficients were compared to the 1970 data of Storm and Israel over the energy interval from 1 to 1000 keV; for elements 10, 20, 30, 40, 50, 60, 70, and 80, the average positive difference between the Storm and Israel coefficients and the coefficients reported here are 1.4%, 2.7%, and 2.6%, for the mass attenuation, mass energy-transfer, and mass-energy absorption coefficients, respectively. The 1969 data compilation of mass attenuation coefficients from McMaster et al. were also compared with the newer LLNL data. Over the energy region from 10 keV to 1000 keV, and from elements Z = 1 to Z = 82 (inclusive), the overall average difference was 1.53% (sigma = 0.85%). While the overall average difference was small, there was larger variation (> 5%) between cross sections for some elements. In addition to coefficient data, other useful data such as the density, atomic weight, K, L1, L2, L3, M, and N edges, and numerous characteristic emission energies are output by the program, depending on a single input variable. The computer source code, written in C, can be accessed and downloaded from the World Wide Web at: http:@www.aip.org/epaps/epaps.html [E-MPHSA-23-1977].
190 citations
TL;DR: The goal of this study was to develop a technique to distinguish benign and malignant breast lesions in secondarily digitized mammograms using six different morphological descriptors, and observed a trend towards increased specificity at low false-negative rates.
Abstract: The goal of this study was to develop a technique to distinguish benign and malignant breast lesions in secondarily digitized mammograms. A set of 51 mammograms (two views/patient) containing lesions of known pathology were evaluated using six different morphological descriptors: circularity, mu R/sigma R (where mu R = mean radial distance of tumor boundary, sigma R = standard deviation); compactness, P2/A (where P = perimeter length of tumor boundary and A = area of the tumor); normalized moment classifier; fractal dimension; and a tumor boundary roughness (TBR) measurement (the number of angles in the tumor boundary with more than one boundary point divided by the total number of angles in the boundary). The lesion was segmented from the surrounding background using an adaptive region growing technique. Ninety-seven percent of the lesions were segmented using this approach. An ROC analysis was performed for each parameter and the results of this analysis were compared to each other and to those obtained from a subjective review by two board-certified radiologists who specialize in mammography. The results of the analysis indicate that all six parameters are diagnostic for malignancy with areas under their ROC curves ranging from 0.759 to 0.928. We observed a trend towards increased specificity at low false-negative rates (0.01 and 0.001) with the TBR measurement. Additionally, the diagnostic accuracy of a classification model based on this parameter was similar to that of the subjective reviewers.
177 citations
TL;DR: A series of tests were designed to investigate the mechanical aspects of DMLC and their implications on dosimetry and create a simple test pattern that may serve as a basis for routine quality assurance checks.
Abstract: It has been shown that intensity-modulated fields have the potential to deliver optimum dose distributions, i.e., high dose uniformity in the target and lower doses in the surrounding critical organs. One way to deliver such fields is by using dynamic multileaf collimation (DMLC). This capability is already available in research mode on some treatment machines. While much effort has been devoted to developing algorithms for DMLC, the mechanical reliability of this new treatment delivery mode has not been fully studied. In this work, we report a series of tests designed to investigate the mechanical aspects of DMLC and their implications on dosimetry. Specifically, these tests were designed to examine (1) the stability of leaf speed, (2) the effect of lateral disequilibrium on dose profiles between adjacent leaves, (3) the significance of acceleration and deceleration of leaf motion, (4) the effect of positional accuracy and rounded-end of the leaves, and (5) create a simple test pattern that may serve as a basis for routine quality assurance checks. Results of these tests are presented. The implications on dosimetry and consideration for the design of leaf motion are discussed.
168 citations
TL;DR: This study indicates that a GA can provide versatility in the design of linear or nonlinear classifiers without a trade-off in the effectiveness of the selected features.
Abstract: We investigated a new approach to feature selection, and demonstrated its application in the task of differentiating regions of interest (ROIs) on mammograms as either mass or normal tissue. The classifier included a genetic algorithm (GA) for image feature selection, and a linear discriminant classifier or a backpropagation neural network (BPN) for formulation of the classifier outputs. The GA‐based feature selection was guided by higher probabilities of survival for fitter combinations of features, where the fitness measure was the area A z under the receiver operating characteristic (ROC) curve. We studied the effect of different GA parameters on classification accuracy, and compared the results to those obtained with stepwise feature selection. The data set used in this study consisted of 168 ROIs containing biopsy‐proven masses and 504 ROIs containing normal tissue. From each ROI, a total of 587 features were extracted, of which 572 were texture features and 15 were morphological features. The GA was trained and tested with several different partitionings of the ROIs into training and testing sets. With the best combination of the GA parameters, the average test A z value using a linear discriminant classifier reached 0.90, as compared to 0.89 for stepwise feature selection. Test A z values with a BPN classifier and a more limited feature pool were 0.90 with GA‐based feature selection, and 0.89 for stepwise feature selection. The use of a GA in tailoring classifiers with specific design characteristics was also discussed. This study indicates that a GA can provide versatility in the design of linear or nonlinear classifiers without a trade‐off in the effectiveness of the selected features.
164 citations
TL;DR: A new method for automatic quantification of the patient setup in three dimensions (3D) using one set of computed tomography (CT) data and two transmission images and was found to be robust for imperfections in the delineation of bony structures in the transmission images.
Abstract: In external beam radiotherapy, conventional analysis of portal images in two dimensions (2D) is limited to verification of in-plane rotations and translations of the patient. We developed and clinically tested a new method for automatic quantification of the patient setup in three dimensions (3D) using one set of computed tomography (CT) data and two transmission images. These transmission images can be either a pair of simulator images or a pair of portal images. Our procedure adjusts the position and orientation of the CT data in order to maximize the distance through bone in the CT data along lines between the focus of the irradiation unit and bony structures in the transmission images. For this purpose, bony features are either automatically detected or manually delineated in the transmission images. The performance of the method was quantified by aligning randomly displaced CT data with transmission images simulated from digitally reconstructed radiographs. In addition, the clinical performance were assessed in a limited number of images of prostate cancer and parotid gland tumor treatments. The complete procedure takes less than 2 min on a 90-MHz Pentium PC. The alignment time is 50 s for portal images and 80 s for simulator images. The accuracy is about 1 mm and 1 degrees. Application to clinical cases demonstrated that the procedure provides essential information for the correction of setup errors in case of large rotations (typically larger than 2 degrees) in the setup. The 3D procedure was found to be robust for imperfections in the delineation of bony structures in the transmission images. Visual verification of the results remains, however, necessary. It can be concluded that our strategy for automatic analysis of patient setup in 3D is accurate and robust. The procedure is relatively fast and reduces the human workload compared with existing techniques for the quantification of patient setup in 3D. In addition, the procedure improves the accuracy of treatment verification in 2D in some cases where rotational deviations in the setup occur.
154 citations
TL;DR: Factors such as its small size, immediate retrieval of results, high accuracy attainable from low applied doses, and as the MOSFET records its dose history make it a suitable in vivo dosimeter where surface and skin doses need to be determined.
Abstract: Radiotherapy x-ray and electron beam surface doses are accurately measurable by use of a MOS-FET detector system. The MOSFET (Metal Oxide Semiconductor Field Effect Transistor) is approximately 200-microns in diameter and consists of a 0.5-microns Al electrode on top of a 1-microns SiO2 and 300-microns Si substrate. Results for % surface dose were within +/- 2% compared to the Attix chamber and within +/- 3% of TLD extrapolation results for normally incident beams. Detectors were compared using different energies, field size, and beam modifying devices such as block trays and wedges. Percentage surface dose for 10 x 10-cm and 40 x 40-cm field size for 6-MV x rays at 100-cm SSD using the MOSFET were 16% and 42% of maximum, respectively. Factors such as its small size, immediate retrieval of results, high accuracy attainable from low applied doses, and as the MOSFET records its dose history make it a suitable in vivo dosimeter where surface and skin doses need to be determined. This can be achieved within part of the first fraction of dose (i.e., only 10 cGy is required.)
TL;DR: In this paper, the authors presented segmentation and classification results of an automated algorithm for the detection of breast masses on digitized mammograms, where potential mass regions were first identified using density-weighted contrast enhancement (DWCE) segmentation applied to single-view mammograms.
Abstract: This paper presents segmentation and classification results of an automated algorithm for the detection of breast masses on digitized mammograms. Potential mass regions were first identified using density-weighted contrast enhancement (DWCE) segmentation applied to single-view mammograms. Once the potential mass regions had been identified, multiresolution texture features extracted from wavelet coefficients were calculated, and linear discriminant analysis (LDA) was used to classify the regions as breast masses or normal tissue. In this article the overall detection results for two independent sets of 84 mammograms used alternately for training and test were evaluated by free-response receiver operating characteristics (FROC) analysis. The test results indicate that this new algorithm produced approximately 4.4 false positive per image at a true positive detection rate of 90% and 2.3 false positives per image at a true positive rate of 80%.
TL;DR: The results indicated that the sensitivity of the improved film was about 40% greater than that of MD-55 film, and the response of theimproved film was found to be uniform within 4% only in one direction of the film.
Abstract: Recently, a new model of radiochromic film has been developed for medical applications to provide a higher sensitivity and better uniformity of response than existing models (i.e., MD-55). Dosimetric characteristics including sensitivity, linearity, reproducibility, uniformity, and dependence on energy and time have been studied experimentally. The characteristics of the new films were compared with those of model MD-55. For these investigations, the two films were exposed to ionizing radiation in the dose range from 1-72 Gy, using gamma-rays from a 60Co teletherapy unit and 6- and 18-MV x rays from a linear accelerator. The response of the exposed film was measured with a helium-neon laser densitometer. The results indicated that the sensitivity of the improved film was about 40% greater than that of MD-55 film. Moreover, the response of the improved film was found to be uniform within 4% only in one direction of the film. The orthogonal direction indicated a nonuniformity of up to 15%, similar to that of model MD-55. Less than 5% energy dependence in the megavoltage photon range was observed for the new film. Complete dosimetric characteristics of the new film are presented.
TL;DR: The transit dosimetry relative dose distribution is found when compared with film and thermoluminescent dosimeter (TLD) measurements and compared with the planning system to agree within 2% in the pelvic region of a humanoid phantom.
Abstract: A method of using electronic portal imaging (EPI) for transit dosimetry is described. In this method, a portal image of the treatment field is first aligned with a digitally reconstructed radiograph (DRR) to geometrically relate the computed tomography (CT) scan, used to generate the DRR, with the EPI. Then the EPI is corrected for scatter within the patient to yield a map of primary fluence striking the detector. This is backprojected through the planning CT data set to yield a distribution of primary fluence within the patient. This distribution is then convolved with dose deposition kemels to yield a map of dose delivery within the patient. Such a distribution may be compared with the dose distribution resulting from the original treatment plan in order to evaluate the adequacy of the treatment. This method has been evaluated using a humanoid phantom. We find the transit dosimetry relative dose distribution when compared with film and thermoluminescent dosimeter (TLD) measurements and compared with our planning system to agree within 2% in the pelvic region of a humanoid phantom.
TL;DR: A compact X-band accelerator mounted on a robotic arm under development for frameless stereotaxic radiosur surgery achieves the same level of targeting precision as conventional frame-based radiosurgery.
Abstract: A compact X‐band accelerator mounted on a robotic arm is under development for frameless stereotaxic radiosurgery. The therapy beam is aimed at the lesion by an imaging system comprised of two diagnostic x‐ray cameras that view the patient during treatment. Patient position and motion are measured by the cameras and communicated in real time to the robotic arm for beam targeting and patient motion tracking. The tests reported here measured the pointing accuracy of the therapy beam and the present targeting and tracking capability of the imaging system. The results show that the system achieves the same level of targeting precision as conventional frame‐based radiosurgery.
TL;DR: The plaque optimization can be performed by dosimetric measurements for the individual patient ("dosimetric treatment planning") and the time consumption for this procedure is less than for a physically correct computer-based therapy planning, e.g., by means of a Monte Carlo simulation.
Abstract: We have developed dosemeters based on plastic scintillators for a variety of applications in radiation therapy. The dosemeters consist basically of a tissue‐substituting scintillator probe, an optical fiber light guide, and a photomultiplier tube. The background light generated in the light guide can be compensated by a simultaneous measurement of the light from a blind fiber. Plastic scintillator dosemeters combine several advantageous properties which render them superior to other dosemeter types for many applications: minimal disturbance of the radiation field because of the homogeneous detector volume and the approximate water equivalence; no dependence on temperature and pressure (under standard clinical conditions) and angle of radiation incidence; no high voltage in the probe; high spatial resolution due to small detector volumes; direct reading of absorbed doses; and a large dynamical range. The high spatial resolution together with direct reading make these detectors suitable for real‐time 3‐D dosimetry using multi‐channel detector systems. Such a system has been developed for eye plaque dosimetry and successfully employed for dosimetric treatment optimization. The plaque optimization can be performed by dosimetricmeasurements for the individual patient (‘‘dosimetric treatment planning’’). The time consumption for this procedure is less than for a physically correct computer‐based therapy planning, e.g., by means of a Monte Carlo simulation.
TL;DR: In this article, the authors investigated a method for positioning the margin required around the clinical target volume (CTV) to account for the random geometrical treatment uncertainties during conformal radiotherapy.
Abstract: In this study we investigate a method for positioning the margin required around the clinical target volume (CTV) to account for the random geometrical treatment uncertainties during conformal radiotherapy. These uncertainties are introduced by patient setup errors and CTV motion within the patient. Three-dimensional dose distributions are calculated for two four-field box techniques and a three-field technique, using rectangular fields. In addition, dose calculations are performed for four prostate cases, treated with a three-field conformal technique. The effects of random rotational and translational deviations on the delivered dose are described as a convolution of the "static" dose with the distribution of the deviations. For the rectangular field techniques, these convolutions are performed with a range of standard deviations (SDs) of the distribution of random translations (0-7 mm in the three directions) and rotations (0 degree-5 degrees around the main axes). Two centers of rotation are considered: the isocenter and a position that is 3.5 cm shifted with respect to the isocenter. For the prostate cases, the random deviations are estimated by combining the results from organ motion and setup accuracy studies. The required margin is defined as the change in the position of the static 95% isodose surface by the convolution and it is approximated by a morphological erosion operator, applied to the static 95% isodose surface. When the center of rotation coincides with the isocenter the change in the position of the static 95% isodose surface can accurately be described by an erosion operator. For the rectangular field techniques, the margin is equal to about 0.7 SD of the distribution of translations, independent of the distribution of rotations. When the center of rotation does not coincide with the isocenter and rotations are considerable, the margin is strongly place dependent, and the accuracy of the approximation by an erosion operator is much lower. In conclusion, margins for random uncertainties can be approximated by a dilation operator (inverse of an erosion operator) when the center of rotational deviations coincides with the isocenter. The size of the margin is about 0.7 SD of the distribution of translations. When rotational deviations are present and the center of rotation does not coincide with the isocenter, the margin can become strongly place dependent and the convolution computation should be incorporated in the planning system.
TL;DR: An iterative convolution/superposition algorithm has been created to reconstruct dose distributions in patients from exit dose measurements during a radiotherapy treatment, based on an extended phantom that includes the patient CT representation and an electronic portal imaging device (EPID).
Abstract: Post-treatment dose verification refers to the process of reconstructing delivered dose distributions internal to a patient from information obtained during the treatment. The exit dose is commonly used to describe the dose beyond the exit surface of the patient from a megavoltage photon beam. Portal imaging provides a method of determining the dose in a plane distal to a patient from a megavoltage therapeutic beam. This exit dose enables reconstruction of the dose distribution from external beam radiation throughout the patient utilizing the convolution/superposition method and an extended phantom. An iterative convolution/superposition algorithm has been created to reconstruct dose distributions in patients from exit dose measurements during a radiotherapy treatment. The method is based on an extended phantom that includes the patient CT representation and an electronic portal imaging device (EPID). The convolution/superposition method computes the dose throughout the extended phantom, which allows the portal dose image to be predicted in the EPID. The process is then reversed to take the portal dose measurement and infer what the dose distribution must have been to produce the measured portal dose. The dose distribution is modeled without knowledge of the incident intensity distribution, and includes the effects of scatter in the computation. The iterative method begins by assuming that the primary energy fluence (PEF) at the portal image plane is equal to the portal dose image, the PEF is then back-projected through the extended phantom and convolved with the dose deposition kernel to determine a new prediction of the portal dose image. The image of the ratio of the computed PEF to the computed portal dose is then multiplied by the measured portal dose image to produce a better representation of the PEF. Successive iterations of this process then converge to the exiting PEF image that would produce the measured portal dose image. Once convergence is established, the dose distribution is determined by back-projecting the PEF and convolving with the dose deposition kernel. The method is accurate, provided the patient representation during treatment is known. The method was used on three phantoms with a photon energy of 6 MV to verify convergence and accuracy of the algorithm. The reconstructed dose volumes agree to within 3% of the forward computation dose volumes. Furthermore, this technique assumes no prior knowledge of the incident fluence and therefore may better represent the dose actually delivered.
TL;DR: It is shown that potentially useful treatment beams can be chosen based on geometric heuristics and that a genetic algorithm can be constructed to find an optimal combination of beams based on a formal objective function.
Abstract: The thesis of this report is that potentially useful treatment beams can be chosen based on geometric heuristics and that a genetic algorithm (GA) can be constructed to find an optimal combination of beams based on a formal objective function. The paper describes the basic principles of a GA and the particular implementation developed. The code represents each plan in the population as two paired lists comprised of beam identifiers and relative weights. Reproduction operators, which mimic sexual reproduction with crossover, mutation, cloning, spontaneous generation, and death, manipulate the lists to grow optimal plans. The necessary gene pool is created by software modules which generate beams, distribute calculation points, obtain clinical constraints, add wedges, and calculate doses. The code has been tested on a set of artificial patients and on four clinical cases: prostate, pancreas, esophagus, and glomus. All demonstrated consistent results, indicating that the code is a reliable optimizer. Additional experiments compared the results for a full set of open beams to the geometrically selected set and the GA code with simulated annealing. Geometric selection of beam directions did not significantly compromise optimization quality. Compared to simulated annealing, the genetic algorithm was equally able to optimize the objective function, and calculations suggest it may be the faster method when the number of beams to be considered exceeds approximately 70.
TL;DR: Four different expressions, derived from the diffusion theory or the random walk model, were used to fit time-resolved reflectance data for the evaluation of tissue optical properties, showing different performances depending on the optical properties of the sample and the experimental conditions.
Abstract: Four different expressions, derived from the diffusion theory or the random walk model, were used to fit time-resolved reflectance data for the evaluation of tissue optical properties. The experimental reflectance curves were obtained from phantoms of known optical parameters (absorption and transport scattering coefficients) covering the range of typical values for biological tissues between 600 and 900 nm. The measurements were performed using an instrumentation for time-correlated single-photon counting. The potential of the four methods in the assessment of the absorption and transport scattering coefficients was evaluated in terms of absolute error, linearity error, and dispersion of data. Each method showed different performances depending on the optical properties of the sample and the experimental conditions. We propose some criteria for the optimal choice of the fitting method to be used in different applications.
TL;DR: Modifications were made to improve the performance of the SIANN and the zero-mean-weight constraint and training-free-zone techniques have been developed and a cross-validation training method was also applied to avoid the overtraining problem.
Abstract: A shift-invariant artificial neutral network (SIANN) has been applied to eliminate the false-positive detections reported by a rule-based computer aided-diagnosis (CAD) scheme developed in our laboratory. Regions of interest (ROIs) were selected around the centers of the rule-based CAD detections and analyzed by the SIANN. In our previous study, background-trend correction and pixel-value normalization were used as the preprocessing of the ROIs prior to the SIANN. A ROI is classified as a positive ROI, if the total number of microcalcifications detected in the ROI is greater than a certain number. In this study, modifications were made to improve the performance of the SIANN. First, the preprocessing is removed because the result of the background-trend correction is affected by the size of ROIs. Second, image-feature analysis is employed to the output of the SIANN in an effort to eliminate some of the false detections by the SIANN. In order to train the SIANN to detect microcalcifications and also to extract image features of microcalcifications, the zero-mean-weight constraint and training-free-zone techniques have been developed. A cross-validation training method was also applied to avoid the overtraining problem. The performance of the SIANN was evaluated by means of ROC analysis using a database of 39 mammograms for training and 50 different mammograms for testing. The analysis yielded an average area under the ROC curve (A(z)) of 0.90 for the testing set. Approximately 62% of false-positive clusters detected by the rule-based scheme were eliminated without any loss of the true-positive clusters by using the improved SIANN with image feature analysis techniques.
TL;DR: The proposed two-step procedure to deal with edge effects in laser-beam transillumination scanning of the human breast was applied to data from breast phantoms and from human breasts, providing higher contrast and detectability in optical mammography with respect to raw data breast images.
Abstract: We have investigated the problem of edge effects in laser-beam transillumination scanning of the human breast. Edge effects arise from tissue thickness variability along the scanned area, and from lateral photon losses through the sides of the breast. Edge effects can be effectively corrected in frequency-domain measurements by employing a two-step procedure: (1) use of the phase information to calculate an effective tissue thickness for each pixel location; (2) application of the knowledge of tissue thickness to calculate an edge-corrected optical image from the ac signal image. The measurements were conducted with a light mammography apparatus (LIMA) designed for feasibility tests in the clinical environment. Operating in the frequency-domain (110 MHz), this instrument performs a transillumination optical scan at two wavelengths (685 and 825 nm). We applied the proposed two-step procedure to data from breast phantoms and from human breasts. The processed images provide higher contrast and detectability in optical mammography with respect to raw data breast images.
TL;DR: The scattered radiation in 6 MV radiotherapy portal images is analyzed and the scatter-to-primary ratio, defined as the ratio of the signal due to scattered radiation to that due to the primary radiation, is defined.
Abstract: The scattered radiation in 6 MV radiotherapy portal images is analyzed. First, a quantity SPR * is studied, by means of Monte Carlo (MC) modeling. SPR * is defined as the ratio, on the central axis, of the signal due to scattered radiation to that due to the primary radiation. The detector model mimics a high‐energy photondetector in the context of transit dosimetry. Second, a physical model of SPR * has been developed from first principles. For a cylindrical phantom, placed symmetrically about the isocenter, it predicts that SPR * depends on the area A at the isocenter of the circular field and the phantom thicknessd as follows. SPR *=k 0 Ad(1+k 1 d)(1+k 2 A), where k 0=0.0266(L 1+L 2)2/(L 1 L 2)2, k 2=−[L −2 1+L −2 2+(L −1 1+L −1 2)2 ((2/3 +(3κ/2))]/2π,L 1 is the source‐to‐isocenter distance, L 2 is the isocenter‐to‐detector distance, and κ is the mean energy of the radiation beam (MeV/0.511). Constant k 1, for which there is no simple expression, depends on L 2. Comparison to the MC data shows that for 60≤L 2≤100 cm the dependence is weak and k 1≅2×10−3 cm−1. The root mean square (rms) agreement between the MC‐derived values of SPR * and the physical model is better than 0.001 over a wide range of A and d values likely to be encountered in clinical practice for L 2≥50 cm. Third, experimental measurements of the scatter‐to‐primary ratio were obtained using our custom built imaging system mounted on a Philips SL25 linear accelerator. In the first experiment, A was varied from 40 to 400 cm2 with L 1=L 2=100 cm with d=20 cm. In the second experiment water depth d was varied from 0 to 28 cm with L 1=L 2=100 cm and A=200 cm2. The rms agreements between the MC data and the experiments were 0.0015 and 0.0045, respectively.
TL;DR: The convolution/superposition method was used to predict the dose throughout an extended volume, which includes a phantom and a portal imaging device, and the ratio of computed primary energy fluence to total dose was found, which enhances the megavoltage portal images by removing scatter blurring.
Abstract: The convolution/superposition method was used to predict the dose throughout an extended volume, which includes a phantom and a portal imaging device. From the calculated dose volume, the dose delivered in the portal image plane was extracted and compared to a portal dose image. This comparison aids in verifying the beam configuration or patient setup after delivery of the radiation. The phantoms used to test the accuracy of this method include a solid water cube, a Nuclear Associates CT phantom, and an Alderson Rando thorax phantom. The dose distribution in the image plane was measured with film and an electronic portal imaging device in each case. The calculated portal dose images were within 4% of the measured images for most voxels in the central portion of the field for all of the extended volumes. The convolution/superposition method also enables the determination of the scatter and primary dose contributions using the particular dose deposition kernels for each contribution. The ratio of primary dose to total dose was used to extract the primary dose from the detected portal image, which enhances the megavoltage portal images by removing scatter blurring. By also predicting the primary energy fluence, we can find the ratio of computed primary energy fluence to total dose. Multiplying this ratio by the measured dose image estimates the relative primary energy fluence at the portal imager. The image of primary energy fluence possesses higher contrast and may be used for further quantitative image processing and dose modeling.
TL;DR: A device that generates low-energy x rays at the tip of a needle-like probe was developed for stereotactic interstitial radiosurgery and to date, 14 patients with metastatic brain tumors have been treated with this device.
Abstract: A device that generates low‐energy x rays at the tip of a needle‐like probe was developed for stereotactic interstitial radiosurgery.Electrons from a small thermionic gun are accelerated to a final energy of up to 40 keV and directed along a 3 mm outside diameter drift tube to a thin Au target, where the beam size is approximately 0.3 mm. All high‐voltage electronics are in the probe housing, connected by low‐voltage cable to a battery‐operated control box. X‐ray output, which is nearly isotropic, consists of a bremsstrahlung spectrum and several lines between 7 and 14 keV, with characteristicradiation contributing 15% of the total energy output. To date, 14 patients with metastatic braintumors have been treated with this device.
TL;DR: While the technique is very sensitive to magnetic field inhomogeneity, stability, and subject motion, it appears to be useful for in vivo temperature change measurement.
Abstract: The temperature sensitivity of the chemical shift of water (approximately 0.01 ppm/degree C) provides a potential method to monitor temperature changes in vivo or in vitro through the changes in phase of a gradient-echo magnetic resonance (MR) image. This relation was studied at 1.5 T in gel materials and in vivo in canine brain and muscle tissue, heated with a radio frequency (rf) annular phased array hyperthermia antenna. The rf fields associated with heating (130 MHz) and imaging (64 MHz) were decoupled using bandpass filters providing isolation in excess of 100 dB, thus allowing simultaneous imaging and rf heating without deterioration of the MR image signal-to-noise ratio. In a gel, temperature sensitivity of the MR image phase was observed to be (4.41 +/- 0.02) phase degrees/degree C for Te = 20 ms, which allowed temperature changes of 0.22 degree C to be resolved for a 50-mm3 region in less than 10 s of data acquisition. In vivo, for Te = 20 ms, the temperature sensitivity was (3.2 +/- 0.1) phase degrees/degree C for brain tissue, (3.1 +/- 0.1) phase degrees/degree C for muscle, and (3.0 +/- 0.2) phase degrees/degree C for a muscle tumor (sarcoma), allowing temperature changes of 0.6 degree C to be resolved in a 16-mm3 volume in less than 10 s of data acquisition. We conclude that, while the technique is very sensitive to magnetic field inhomogeneity, stability, and subject motion, it appears to be useful for in vivo temperature change measurement.
TL;DR: A miniature, battery operated 40 kV x-ray device has been developed for the interstitial treatment of small tumors in humans, and behaves essentially as a point isotropic source with an effective energy of 20 keV at a depth of 10 mm in water.
Abstract: A miniature, battery operated 40 kV x-ray device has been developed for the interstitial treatment of small tumors ( < 3 cm diam) in humans. X rays are emitted from the tip of a 10 cm long, 3 mm diameter probe that is stereotactically inserted into the tumor. The beam, characterized by half-value layer (HVL), spectrum analysis, and isodose contours, behaves essentially as a point isotropic source with an effective energy of 20 keV at a depth of 10 mm in water. The absolute output from the device was measured using a parallel plate ionization chamber, modified with a platinum aperture. The dose rate in water determined from these chamber measurements was found to be nominally 150 cGy/min at a distance of 10 mm for a beam current of 40 microA and voltage of 40 kV. The dose in water falls off approximately as the third power of the distance. To date, 14 patients have been treated with this device in a phase I clinical trial.
TL;DR: The GA application provides an automated design strategy for prostate implant planning, and at the same time affords the potential for systematic optimization of a set of end points that can sustain practical variations.
Abstract: A genetic algorithm (GA) is presented for the optimization of template- and ultrasound-guided prostate implants. The end points for optimization are incorporated in an objective function of separable cardinal utility terms. As an application of the GA, the minimum 103Pd total source strength required to deliver a given dose was correlated with the average dimension for prostate implants carried out under the current template and seed spacing protocols. Significant improvements in quality were observed, in terms of both the minimum peripheral dose and tumor cell surviving fractions, when GA-optimized implants were compared to the corresponding unoptimized implants for given target volumes. In addition, numerical simulation of source displacements indicates that the dosimetric and radiobiologic advantages of GA optimization can tolerate a reasonable level of seed placement uncertainties observed clinically. In summary, the GA application provides an automated design strategy for prostate implant planning, and at the same time affords the potential for systematic optimization of a set of end points that can sustain practical variations.
TL;DR: The Monte Carlo-based algorithm presented in this article produces accurate absorbed dose distributions due to patient-specific density and radionuclide activity distributions that were investigated to investigate the implications of replacing a mean absorbed dose with a distribution and of neglecting atomic number and density variations for various patient geometries and energies.
Abstract: In internal emitter therapy, an accurate description of the absorbed dose distribution is necessary to establish an administered dose-response relationship, as well as to avoid critical organ toxicity. Given a spatial distribution of cumulated activity, an absorbed dose distribution that accounts for the effects of attenuation and scatter can be obtained using a Monte Carlo method that simulates particle transport across the various densities and atomic numbers encountered in the human body. Patient-specific information can be obtained from CT and SPECT or PET imaging. Since the data from these imaging modalities is discrete, it is necessary to develop a technique to efficiently transport particles across discrete media. The Monte Carlo-based algorithm presented in this article produces accurate absorbed dose distributions due to patient-specific density and radionuclide activity distributions. The method was verified by creating CT and SPECT arrays for the Medical Internal Radionuclide Dose (MIRD) Committee's Standard Man phantom, and reproducing the spatially averaged specific absorbed fractions reported in MIRD Pamphlet 5. The algorithm was used to investigate the implications of replacing a mean absorbed dose with a distribution, and of neglecting atomic number and density variations for various patient geometries and energies. For example, the I-131 specific absorbed fraction for spleen to liver is the same as for liver to spleen, yet the distributions were different. Furthermore, neglecting atomic number variations across the vertebral bone led to an overestimation of I-125 absorbed dose by an order of magnitude, while no error was observed for I-131.
TL;DR: For electron beam reference dosimetry in radiotherapy, it is shown that by choosing the reference depth as dref = 0.6R(50)-0.1 cm, where R50 is the half-value depth in centimeters, the Spencer-Attix water-to-air stopping-power ratio at dref is given by (Llp)airw.
Abstract: For electron beam reference dosimetry in radiotherapy, it is shown that by choosing the reference depth as dref = 0.6R(50)-0.1 cm, where R50 is the half-value depth in centimeters, the Spencer-Attix water-to-air stopping-power ratio at dref is given by (Llp)airw = 1.2534 - 0.1487 (R50)0.2144. This is derived from data for (Llp)airw obtained from realistic Monte Carlo simulations for 24 clinical beams. The rms deviation of this expression from the Monte Carlo calculations is 0.16%, with a maximum deviation of 0.26%. This approach fully takes into account the spectral differences between real electron beams of the same R50 and allows an absorbed-dose calibration at a standards laboratory to be easily and accurately transferred to a reference clinical beam. Using a single parameter to specify (Llp)airw, rather than the two parameters (R50 and depth) needed when the reference depth is chosen as the depth of dose maximum, has the potential to greatly simplify electron beam dosimetry protocols and allows the use of a similar formalism for photon and electron beam dosimetry. For use in converting a depth-ionization curve into a depth-dose curve, a somewhat less accurate but general expression for (Llp)w(air) as a function of R50 and depth is presented.