摘要:

Receiver operating characteristic (ROC) methodology is widely used in evaluating medical imaging modalities. While appropriate in some cases, it has several drawbacks when the detection task, e.g., nodule detection, involves localizing the abnormality. Free‐response receiver operating characteristic (FROC) methodology offers a more natural framework to describe observer performance in such studies and has other advantages. Due to the lack of a statistical analysis procedure comparable to the maximum likelihood procedure (rocfitprogram) available for ROC studies, the FROC method has not gained widespread acceptance. This work presents and solves a two parameter model for the statistical analysis of FROC data. The model assumes that the probability density of the signal stimuli is normally distributed, as is the probability density for producing one or more false positives per image. A program (frocfit) is described for estimating the parameters and their uncertainties from experimental data. An index of performance is proposed to quantify observer performance in FROC experiments. Application of this methodology to several FROC data sets produced good to excellent fits.

ISSN:0269-3127    

DOI:10.1118/1.596358

出版商:American Association of Physicists in Medicine    

年代:1998

数据来源: WILEY

摘要:

Studying motion in the gastrointestinal (G.I.) tract is the first step towards our goal of designing a digital algorithm for real‐time noise and motion blurring reduction by temporal and spatial averaging in x‐ray fluoroscopy. The present work concerns the types of motion relevant to G.I. fluoroscopy (e.g., motion of the walls of the oesophagus, the stomach, the small intestine, and the colon), and measurement of the range of velocities in different parts of the G.I. tract from tape recorded fluoroscopic sequences. The relationship between image contrast and velocity, the effects of motion on system response, and in particular the influence of these parameters on the evaluation of velocities will also be discussed.

ISSN:0269-3127    

DOI:10.1118/1.596359

出版商:American Association of Physicists in Medicine    

年代:1998

数据来源: WILEY

摘要:

A method for photon beam dose calculations is described. The primary photon beam is raytraced through the patient, and the distribution of total radiant energy released into the patient is calculated. Polyenergetic energy deposition kernels are calculated from the spectrum of the beam, using a database of monoenergetic kernels. It is shown that the polyenergetic kernels can be analytically described with high precision by {Aexp(−ar)+Bexp(−br)}/r2, whereA,a,B, andbdepend on the angle with respect to the impinging photons and the accelerating potential, andris the radial distance. Numerical values ofA,a,B, andbare derived and used to convolve energy deposition kernels with the total energy released per unit mass (TERMA) to yield dose distributions. The convolution is facilitated by the introduction of the collapsed cone approximation. In this approximation, all energy released into coaxial cones of equal solid angle, from volume elements on the cone axis, is rectilinearly transported, attenuated, and deposited in elements on the axis. Scaling of the kernels is implicitly done during the convolution procedure to fully account for inhomogeneities present in the irradiated volume. The number of computational operations needed to compute the dose with the method is proportional to the number of calculation points. The method is tested for five accelerating potentials; 4, 6, 10, 15, and 24 MV, and applied to two geometries; one is a stack of slabs of tissue media, and the other is a mediastinum‐like phantom of cork and water. In these geometries, the EGS4 Monte Carlo system has been used to generate reference dose distributions with which the dose computed with the collapsed cone convolution method is compared. Generally, the agreement between the methods is excellent. Deviations are observed in situations of lateral charged particle disequilibrium in low density media, however, but the result is superior compared to that of the generalized Batho method.

ISSN:0269-3127    

DOI:10.1118/1.596360

出版商:American Association of Physicists in Medicine    

年代:1998

数据来源: WILEY

摘要:

Dose distributions in the cranium with fixed and with variably modulated proton beams were compared. The variable modulation was designed to tailor the proximal high‐dose region of each field to the target volume surface whereas the fixed modulation beams had a constant modulation determined by the greatest extent of the target. Dose‐volume histograms of normal tissues were compared, as were the estimated complication probabilities. Twelve patients with chordomas or chondrosarcomas of the base of skull who had been treated to approximately 70 cobalt Gray equivalent (CGE) were evaluated. Dose distributions of the actual treatments were compared to those which would have been delivered had the proton beams been variably modulated; two patients for whom x‐ray plans were available were also evaluated. The greatest difference in dose between the variable and fixed modulation proton beams, averaged over all the patients, was 13.8 CGE (8.0–18.0 CGE range). Much of this reduction occurred in the brain, particularly the temporal lobes. In those temporal lobes receiving significant doses, variable modulation reduced the volume receiving more than 54 CGE by 3.0 cc; all temporal lobes had at least a 5 CGE difference to some portion, half had more than 10 CGE and three more than 15 CGE difference to some portion. The optic structures, brainstem and spinal cord received from 1 to 3 CGE less dose with the variability modulated beams. Eight of the parotid glands received more than 20 CGE to more than half their volume with the fixed modulation beams; in these, variable modulation reduced the mean dose by 5.3 CGE. The reduction in integral dose with variable as compared to fixed modulation was in the range 3 to 12%; this gain was considerably less than the gain for uniformly modulated proton beams over x‐rays in the two patients for whom x‐ray plans were available.

ISSN:0269-3127    

DOI:10.1118/1.596428

出版商:American Association of Physicists in Medicine    

年代:1998

数据来源: WILEY

摘要:

An approach to photodynamic treatment of tumors is the interstitial implantation of fiber optic light sources. Dosimetry is critical in identifying regions of low light intensity in the tumor which may prevent tumor cure. We describe a numerical technique for calculating light distributions within tumors, from multiple fiber optic sources. The method was tested using four translucent plastic needles, which were placed in a 0.94×0.94 cm grid pattern within excised Dunning R3327‐AT rat prostate tumors. A cylindrical diffusing fiber tip, illuminated by 630 nm dye laser light was placed within one needle and a miniature light detector was placed within another. The average penetration depth in the tumor region between the two needles was calculated from the optical power measured by the detector, using a modified diffusion theory. Repeating the procedure for each pair of needles revealed significant variations in penetration depth within individual tumors. Average values of penetration depth, absorption coefficient, scattering coefficient, and mean scattering cosine were 0.282 cm, 0.469 cm−1, 250 cm−1and 0.964, respectively. Calculated light distributions from four cylindrical sources in tumors gave reasonable agreement with direct light measurements using fiber optic probes.

ISSN:0269-3127    

DOI:10.1118/1.596361

出版商:American Association of Physicists in Medicine    

年代:1998

数据来源: WILEY

摘要:

Changes in dose distributions in buildup region resulting from the presence of lead, aluminum, and lucite absorbers above the surface of a polystyrene phantom were evaluated. The surface dose, as a function of the absorber thickness, is presented as well as the influence of the air gap between the lead absorber and the phantom surface. It has been found that the surface dose does not depend on absorber thickness for absorbers thicker than the range of secondary electrons in the absorber material (after corrections for the attenuation of the primary beam in the absorber). Similarly, the depth dose curves in the phantom were elevated only at depths lower than the range of secondary electrons in the phantom. The applicability of the presented data in clinical radiotherapy is discussed.

ISSN:0269-3127    

DOI:10.1118/1.596362

出版商:American Association of Physicists in Medicine    

年代:1998

数据来源: WILEY

摘要:

Testing computer‐controlled linear accelerators for patient safety and proper patient dose delivery requires that certain beam characteristics be monitored over an extended period of time. Computer‐controlled conformal radiation therapy using asymmetric collimator jaw settings necessitates stable symmetric treatment beams. Long term beam symmetry measurements have been performed on a Philips SL20 dual energy computer‐controlled linear accelerator. Symmetry in both the radial and transverse axis of each x‐ray beam was monitored for eight gantry positions. These measurements were undertaken to determine the effectiveness of the SL20 beam steering system during dose delivery of 50 monitor units (MU) per field. Evaluation of the data shows that careful beam steering setup procedures result in x‐ray beams in which fluctuations in symmetry as a function of gantry angle are within ±1.5%. Day to day instabilities produce a total overall variation in beam symmetry on the order of ±2.0%. Results suggest the measurement of symmetry as a function of gantry position become a routine quality assurance procedure for this accelerator.

ISSN:0269-3127    

DOI:10.1118/1.596363

出版商:American Association of Physicists in Medicine    

年代:1998

数据来源: WILEY

摘要:

In this study the ultrasonic field distortions and the temperature measurement errors caused by temperature sensing probes were investigated. It was found that probes with diameters equal to or greater than 1/2 of the square root of the wavelength (λ) scatter and reflect the waves, and thus distort the field significantly. Smaller probes down to λ1/2/5 in diameter had a detectable effect which was very local and, therefore, will probably not have a significant effect on the overall temperature distribution. When the temperature measurement errors were studied, even the smallest probes showed some self‐heating artifact and its magnitude depended on the probe size, material, structure, orientation, and the operating frequency. This error was small with most of the nonultrasound absorbing probes (such as bare wires or stainless steel needle probes), but significant with plastic shielded thermocouples. The energy absorption associated with plastic coated probes can be reduced by orienting the probe parallel to the beam, by scanning the focus in such a manner that the focus is not passing on the sensor, and by inserting the probe in a stainless steel catheter. The temperature measurement error can also be reduced by covering the plastic probe with a high‐ (or low‐) acoustical impedance material around the sensor. This would scatter the sound around the probe, thus preventing energy absorption in the plastic. This arrangement can preserve the desirable properties of plastic probes (mainly flexibility, strength, and electrical isolation from the patient) while at the same time allowing fairly absorption artifact free measurements.

ISSN:0269-3127    

DOI:10.1118/1.596364

出版商:American Association of Physicists in Medicine    

年代:1998

数据来源: WILEY

摘要:

The Monte Carlo method was used to study perturbations of single I‐125 seed dose distributions created by the presence of one or three neighboring seeds for the case of seeds immersed in a water phantom. Perturbation factors were determined within the geometric shadow of neighboring seeds for two‐seed designs, four‐seed spacings, and several choices of dose point. The results were compared to dose estimates obtained by the simple superposition of single‐seed data for one‐ and two‐plane implants. Some significant differences were found.

ISSN:0269-3127    

DOI:10.1118/1.596320

出版商:American Association of Physicists in Medicine    

年代:1998

数据来源: WILEY

摘要:

For interstitial implants with radioactive Au‐198 seeds, the dosimetric calculations usually ignore the finite source size and employ a point source approximation, resulting in a computed isotropic dose distribution. However, the measured radiation fluence from Au‐198 seeds is reported to be anisotropic, suggesting some amount of dose anisotropy in tissue. We have measured this dose anisotropy around Au‐198 seeds (2.5×0.8 mm2) using Kodak X‐OMAT XV‐2 films placed in contact with individual seeds in a phantom. Autoradiographs obtained for various exposure times were digitized and studied with an image analysis computer network. The network's overall spatial resolution was about 0.01 cm. The optical isodensity contours around the seed were obtained. The useful range of optical density (OD) for the system was found to be 1 to 2 OD units. Within this range, the shape of an optical isodensity contour would be identical to an isodose contour. Proximal to the source, the contours were elliptical in shape, elongated along the seed axis. However, further away, the elongation was in the direction normal to the seed axis. This was in agreement with the reported data on the radiation fluence around Au‐198 seeds. It was concluded that measurable dose anisotropy existed around Au‐198 seeds. However, it was too small to be of any clinical significance.

ISSN:0269-3127    

DOI:10.1118/1.596321

出版商:American Association of Physicists in Medicine    

年代:1998

数据来源: WILEY