摘要:

An important complication of osteoporosis is fracture. Alteration in bone structure, as well as decreased bone mass, contribute to the tendency to fracture in osteoporosis. Current methods that measure bone mass alone show substantial overlap of the measurements of osteoporotic patients who fracture with those that do not. Our aim is to develop noninvasive methods of evaluating bone structure on plain film radiographs to better predict fracture risk in osteoporosis. Regions of interest (ROIs) were selected from digitized lateral lumbar spine radiographs of 43 patients being seen in an osteoporosis clinic. The fractal dimension of these ROIs was estimated using a surface area method. The ability of fractal dimension to distinguish between cases that had fracture elsewhere in the spine from those that did not, was evaluated using receiver operating characteristic (ROC) analysis. These results were compared with ROC analysis for these same patients using bone mineral density (BMD) measurements (bone mass). Significantly largerAz(area under ROC curve) values were obtained using fractal dimension (0.87) than from using BMD (0.58), indicating a better test performance using fractal dimension. Therefore, computerized radiographic methods to evaluate bone structure, such as fractal analysis, may be helpful in better determining fracture risk in osteoporosis.

ISSN:0094-2405    

DOI:10.1118/1.597390

出版商:American Association of Physicists in Medicine    

年代:1998

数据来源: WILEY

摘要:

A digital chest radiography system has been developed, with a detector based on the photoelectric properties of amorphous selenium. The selenium layer is deposited on a cylindrical aluminium drum, large enough to cover the full field of view for chest imaging. The electrostatic charge image which is formed on the selenium surface after x‐ray exposure is read out by electrometer probes using fast drum rotation. For a physical evaluation of the attainable image quality, the characteristic curve, the modulation transfer function, and the noise spectra were measured. From these measurements, the signal‐to‐noise properties of the detector in terms of detective quantum efficiency (DQE) and noise equivalent quanta (NEQ) were derived. The results show that the selenium‐based detector has a wide dynamic range and a significantly better DQE than screen‐film and storage phosphor systems for spatial frequencies below the Nyquist limit (2.7 lp/mm). As a consequence, the detectability of small, low‐contrast details is considerably improved.

ISSN:0094-2405    

DOI:10.1118/1.597389

出版商:American Association of Physicists in Medicine    

年代:1998

数据来源: WILEY

摘要:

A computer‐aided diagnosis (CAD) scheme has been developed in our laboratory for the detection of clustered microcalcifications in digital mammograms. In this study, we apply a shift‐invariant neural network to eliminate false‐positive detections reported by the CAD scheme. The shift‐invariant neural network is a multilayer back‐propagation neural network with local, shift‐invariant interconnections. The advantage of the shift‐invariant neural network is that the result of the network is not dependent on the locations of the clustered microcalcifications in the input layer. The neural network is trained to detect each individual microcalcification in a given region of interest (ROI) reported by the CAD scheme. A ROI is classified as a positive ROI if the total number of microcalcifications detected in the ROI is greater than a certain number. The performance of the shift‐invariant neural network was evaluated by means of ajackknife(orholdout) method and ROC analysis using a database of 168 ROIs, as reported by the CAD scheme when applied to 34 mammograms. The analysis yielded an average area under the ROC curve (Az) of 0.91. Approximately 55% of false‐positive ROIs were eliminated without any loss of the true‐positive ROIs. The result is considerably better than that obtained in our previous study using a conventional three‐layer, feed‐forward neural network. The effect of the network structure on the performance of the shift‐invariant neural network is also studied.

ISSN:0094-2405    

DOI:10.1118/1.597177

出版商:American Association of Physicists in Medicine    

年代:1998

数据来源: WILEY

摘要:

Recognition of the importance of performing simulated screen‐light sensitometry of medical x‐ray films for the purpose of processor quality control has increased over the past several years. As a result there is a greater need to provide new techniques for performing simulated screen‐light sensitometry. Medical films with reduced intensifying screen‐light crossover intended to achieve reduced blur and higher spatial resolution pose particular problems in doing simulated screen‐light sensitometry if care is not taken to choose a proper simulated light sensitometer with the capability of simultaneous double‐sided exposures. Misleading and incorrect sensitometric data can be obtained for film contrast evaluation if a single side exposure is used. Asymmetric, near‐zero crossover films pose even greater problems as proper orientation of the film and proper degree of light output asymmetry need be achieved in order to obtain correct sensitometry. The films used in this study were three double emulsion films varying in crossover from 3% to 24%. Of the two very‐low‐crossover films, one had symmetric emulsion layers while the second featured emulsion layers which were asymmetric in terms of contrast and speed. Sensitometric data show several curve shapes with significant distortions, depending on orientation, for the asymmetric, low‐crossover film when exposed using a single‐sided exposure. Only by using a double‐sided exposure and an appropriate neutral density filter to simulate the degree of screen‐light asymmetry in this system could one achieve a characteristic curve comparable to that achieved by inverse square sensitometry.

ISSN:0094-2405    

DOI:10.1118/1.597168

出版商:American Association of Physicists in Medicine    

年代:1998

数据来源: WILEY

摘要:

A feasibility study is presented for a small, low‐cost, dedicated device for positron emission mammography. Two detector arrays above and below the breast would be placed in a conventional mammography unit. These detectors are sensitive to positron annihilation radiation, and are connected to a coincidence circuit and a multiplane image memory. Images of the distribution of positron‐emitting isotope are obtained in real time by incrementing the memory location at the intersection of each line of response. Monte Carlo simulations of a breast phantom are compared with actual scans of this phantom in a conventional PET scanner. The simulations and experimental data are used to predict the performance of the proposed system. Spatial resolution experiments using very narrow bismuth germanate BGO crystals suggest that spatial resolutions of about 2 mm should be possible. The efficiency of the proposed device is about ten times that of a conventional brain scanner. The scatter fraction is greater, but the scattered radiation has a very flat distribution. By designing the device to fit in an existing mammography unit, conventional mammograms can be taken after the injection of the radio‐pharmaceutical allowing exact registration of the emission and conventional mammographic images.

ISSN:0094-2405    

DOI:10.1118/1.597169

出版商:American Association of Physicists in Medicine    

年代:1998

数据来源: WILEY

摘要:

A general expression is derived for the NMR signal from a fluid undergoing random directional flow such as encountered within the microcirculation. The dependence of the echo amplitude on flow velocity, sample morphology, and experimental parameters are described in terms of a temporal velocity autocorrelation function. The width of the correlation function determines whether the flow can properly be described as diffusive. Comparison is made between the velocity autocorrelation method outlined here and the IVIM model for tissue perfusion. Conditions for the validity of the latter approach for extracting physiologic information from apparent diffusion measurements are discussed. The approach outlined leads to a more robust measure of microcirculatory blood velocity from NMR measurements.

ISSN:0094-2405    

DOI:10.1118/1.597170

出版商:American Association of Physicists in Medicine    

年代:1998

数据来源: WILEY

摘要:

The electrical conductivity and relative permittivity of malignant and normal human tissues were measured at frequencies from 50 to 900 MHz. The measurements were made between 23 and 25 °C using a network analyzer connected to a flat‐ended coaxial probe that was pressed against the freshly excised tissue samples. The malignant tissues were of the following normal tissue origin: bladder, colon, kidney, liver, lung, lymph nodes, mammary gland, spleen, and testes. The normal tissues included: colon, kidney, liver, lung, mammary gland, and muscle. Normal tissue samples of bladder, lymph, spleen, and testes were not available. In general, at all frequencies tested, both conductivity and relative permittivity were greater in malignant tissue than in normal tissue of the same type. For tissues of the same type, the differences in electrical properties from normal to malignant were the least for kidney (about 6% and 4% average differences over the frequency range in permittivity and conductivity, respectively), and these differences were the greatest for mammary gland (about 233% and 577% average differences in permittivity and conductivity, respectively). To illustrate a potential use of these data in hyperthermia applications, frequency‐selective heating of malignant tissue (modeled as a sphere) surrounded by host normal tissue is calculated from the measured electrical properties for certain tissues.

ISSN:0094-2405    

DOI:10.1118/1.597312

出版商:American Association of Physicists in Medicine    

年代:1998

数据来源: WILEY

摘要:

A method to register pairs of portal images quickly and accurately has been developed. The approach uses a cross‐correlation operator to find the optimal match between corresponding anatomic regions that have been selected by a user on pairs of portal images. The cross‐correlation operator determines the translation that best registers each pair of anatomic features independently, and then the images are translated, rotated, and scaled so that the least squares difference between the coordinates of all of the paired regions is minimized. Tests using simulated images have shown that the accuracy of the algorithm is dependent on (i) the size and shape of the structures within the paired regions; (ii) the subject contrast of the anatomic features being matched; (iii) the rotational difference between the anatomic features; and, (iv) the noise in the image. Tests using contrast‐detail phantom images have shown that the semiautomated registration algorithm is only slightly less accurate than human observers, but is considerably faster. The semiautomated algorithm shows great promise as a method of quickly and accurately estimating discrepancies in patient positioning when used in conjunction with an on‐line portal imaging system.

ISSN:0094-2405    

DOI:10.1118/1.597391

出版商:American Association of Physicists in Medicine    

年代:1998

数据来源: WILEY

摘要:

This work evaluates the application of AAPM task group 25 (TG25) methodology for determination of central axis depth dose for a radiotherapy linear accelerator, whose dual scattering foil system and applicators were recently modified. The percent depth dose (%DD) and the dose output factor have been measured for square and rectangular fields at 100‐ and 110‐cm source‐to‐surface distance (SSDs). At 100‐cm SSD, results showed that %DD for a specific energy and field size can vary with applicator, the largest variation being for the 20‐MeV, 10×10‐cm field where a spread of ±2.5% or ±3 mm about the mean %DD is observed. The square‐root method determines rectangular field %DD within 1%. Output factors for rectangular fields are calculated from square field values more accurately using a square‐root method than the equivalent‐square method recommended by TG25. At 110‐cm SSD, the %DD calculated from that at 100‐cm SSD using an inverse square factor does not agree with measured values for all fields. The maximum difference observed for the 20‐MeV, 6×6‐cm field was 5.5% or 10 mm. Output data at the 110‐cm SSD show that the square‐root method is suitable for determination of the air‐gap correction factors of rectangular fields. In summary, the recommendations of TG25 work reasonably well for central axis electron beam dosimetry for this version of a radiotherapy linear accelerator, except in limited cases where applicator‐scattered electrons apparently cause minor but clinically significant discrepancies.

ISSN:0094-2405    

DOI:10.1118/1.597313

出版商:American Association of Physicists in Medicine    

年代:1998

数据来源: WILEY

摘要:

A new type of direct reading semiconductor dosimeter has been investigated as a radiation detector for photon and electron therapy beams of various energies. The operation of this device is based on the measurement of the threshold voltage shift in a custom‐built metal oxide‐silicon semiconductor field effect transistor (MOSFET). This voltage is a linear function of absorbed dose. The extent of the linearity region is dependent on the voltage controlled operation during irradiation. Operating two MOSFETS at two different biases simultaneously during irradiation will result in sensitivity (V/Gy) reproducibility better than ±3% over a range in dose of 100 Gy and at a dose per fraction greater than 20×10−2Gy. The modes of operation give this device many advantages, such as continuous monitoring during irradiation, immediate reading, and permanent storage of total dose after irradiation. The availability and ease of use of these MOSFET detectors make them very promising in clinical dosimetry.

ISSN:0094-2405    

DOI:10.1118/1.597314

出版商:American Association of Physicists in Medicine    

年代:1998

数据来源: WILEY