Search PubMed⌕ Search

Biomedical subjects

D J Drost

Publications and source records attributed to D J Drost.

42 records · Page 3Linked to original sources
PubMedDate not supplied

A xenon ionization detector for scanned projection radiography: theoretical considerations.

Xenon ionization detectors have been used successfully in computed tomography (CT) scanners; however, the detector design used in CT scanners does not provide sufficient spatial resolution for scanned projection radiography. We have been investigating a new design of xenon detector with individual element widths of 0.5 mm. In this design, there are no metal septa separating individual elements. As a result, detection efficiency is better than the design with septa, and construction of an array with submillimeter element widths is simpler; however, crosstalk will now occur between elements. Theoretical calculations of efficiency and resolution for our septaless design of xenon detector are presented. Results of these calculations indicate that for a spectrum of 100 kVp, element dimensions of 0.5 mm X 0.5 mm X 10 cm, a front window of 0.5-mm aluminum, and a xenon pressure of 20 atm, quantum efficiency will be greater than 95%, detective quantum efficiency (DQE) will be approximately 75%, and both energy and conversion efficiencies will be limited to 50% by K-fluorescent escape. The calculations also predict that for the same design, the reduction in lesion contrast induced by crosstalk will be less than 10% for all typical spectra. These theoretical results have encouraged us to pursue the construction of a prototype septaless xenon detector for scanned projection radiography.

Models, Theoretical↗
PubMedDate not supplied

A xenon ionization detector for scanned projection radiography: 95-channel prototype evaluation.

We have designed, constructed, and tested a 95-channel prototype xenon ionization detector for use in scanned projection radiography (SPR). This detector has higher spatial resolution, is more dose efficient, and is easier to construct than computed tomography (CT) xenon ionization detectors. It consists of two parallel plates separated by a 2-mm gap filled with xenon gas at 2 MPa (20 atm). One plate is a high-voltage electrode while the other is a circuit board etched to form an array of metal collector strips focused on the x-ray source. The resulting detector channels are 0.5 mm wide and 6 cm long. In this paper we present results from measurements of system noise and detector channel calibration. We compared the detector system to a screen/film system and found that it allows the detection of structures with 0.17% radiographic contrast compared to 2% contrast required for detection with screen/film when tested by imaging a 10-cm-thick Lucite phantom with a 10 X 10(-6) C/kg exposure. From images of resolution test patterns, the limiting resolution of the detector is 2.0 1p/mm at 1.6 magnification. Images of reduction mammoplasty tissue samples, obtained with 1/17 the exposure of screen/film images, had the same low-contrast sensitivity but contained less high-contrast detail than the film images.

Radiographic Image Enhancement↗
PubMedDate not supplied

A xenon ionization detector for scanned projection radiography: Xenon/Freon 13B1 comparison.

In a companion paper, we reported on the construction and testing of a xenon gas ionization detector for use in line scanned projection radiography. Experimental results indicate that the detector has sufficient resolution for chest radiography, but higher resolution is required for mammography. Theoretical analysis suggested that a detector pressurized with Freon 13B1 would have better resolution and higher x-ray energy efficiency than a xenon-filled detector for energies below 60 keV. In this paper we compared, theoretically and experimentally, Freon 13B1 to xenon as the detector gas. For a 120-kVp x-ray spectrum, the detector filled with 2.0 MPa of xenon had less channel-to-channel crosstalk, a higher quantum efficiency (QE), and twice the output signal than the detector filled with 1.4 MPa of Freon (highest possible pressure at room temperature), while for a 60-kVp spectrum, crosstalk is the same, but the detector has slightly higher QE and 1.4 times the energy efficiency when filled with Freon instead of xenon. We conclude that xenon is better for high-kVp imaging, while Freon at a lower pressure is slightly better for low-kVp imaging.

Bromochlorofluorocarbons↗
PubMedDate not supplied

Dual-energy mammography: initial experimental results.

Dual-energy x-ray techniques may be able to enhance the detectability of calcifications in mammographic examinations by removing the background "clutter" caused by contrast between adipose and glandular tissue. This hypothesis is examined experimentally by implementation of dual-energy imaging on a prototype digital scanned projection radiography system developed in our laboratory. A model of the propagation of signal and noise in dual-energy processing for a given radiation dose is validated by measurements from phantom images. The experimental imaging system has low spatial resolution and cannot be operated at dose-optimum energies; however, since both the single- and dual-energy images are subject to the same technical limitations, a comparison of such images allows an assessment of the benefits of dual energy. Experimental images of breast tissue specimens, showing improved detectability of calcifications when obscuring background clutter is removed, are presented. The dose required for a given signal-to-noise ratio can be reduced by smoothing the higher energy image prior to dual-energy processing. For practical implementation, it is reasonable to smooth the higher energy image such that its variance is reduced fourfold.

Computers↗
PubMedDate not supplied

A cosine modulation artifact in modulation transfer function computations caused by the misregistration of line spread profiles.

Modulation transfer functions (MTFs) are used to analyze the spatial frequency transfer characteristics of medical imaging systems. By definition, accurate MTFs should not include the effects of image noise and they should not be aliased. Therefore, many techniques used to compute MTFs register and average together multiple profiles to improve both signal to noise and/or eliminate aliasing. It is demonstrated that improper registration of individual profiles can cause errors of up to 100% in the MTF. Computer modelling shows that a maximum allowable error of 2% in the MTF requires a registration precision of +/- 1/9 of a pixel for each profile, if the profile was sampled at twice the cutoff frequency of the MTF. One suggested registration method, demonstrated with experimental magnetic resonance image data, is 1.5 times more accurate than the minimum requirement.

Algorithms↗