Chris C. Shaw

FRACTAL ANALYSIS OF TRABECULAR PATTERNS IN PROJECTION RADIOGRAPHS : AN ASSESSMENT

RATIONALE AND OBJECTIVES.Fractal analysis of digitized images has been investigated in recent years as a potential measure of structural bone strength. Several technical issues associated with such measurements are assessed. METHODS.In a series of experiments using a hand phantom, the effects of system noise and modulation transfer function on fractal dimension were explored. The authors evaluated a method for correcting the estimated power spectrum using a step-wedge image exposed and digitized under identical conditions as a reference. RESULTS.System noise and modulation transfer function significantly affect estimated fractal dimension in bony regions computed from conventional radiographs. CONCLUSIONS.Before conventional radiographs are used for fractal analysis in the clinical environment, many of the technical problems associated with this methodology must be addressed.

Comparison of three different techniques for dual-energy subtraction imaging in digital radiography: A signal-to-noise analysis

Dual-energy subtraction imaging techniques allow the tissue and bone structures in the patient to be visualized and studied in two separate images, thus removing the obscurity associated with overlapping of the two structures. In addition, they allow the subtraction image signals to be used for quantifying the tissue and bone thicknesses. Thus, capability for dual-energy subtraction imaging is often incorporated with new digital radiography systems. There are three different approaches to dual-energy image subtraction imaging techniques. Among them, the dual-kilovolt (peak) [kV(p)] and sandwich detector techniques have been two widely used approaches. A third approach is the single-kV(p) dual-filter technique, which allows some flexible control of the spectra while avoiding the technical complexity of kV(p) value switching in slitscan imaging. In this report, the noise properties associated with these three techniques are studied and compared by computing the noise variances in the subtraction image signals as a function of the kV(p) values and filter thicknesses. It was found that the dual-kVp technique results in the least noisy subtraction images, whereas the dual-filter technique results in slightly less noisy subtraction images than the sandwich detector technique. Following optimization of the kV(p) value and filter thicknesses, the dual-filter and sandwich detector techniques result in a noise level of approximately three and four times higher than that resulted from the dual-kV(p) technique, respectively.

Computed radiography versus screen-film mammography in detection of simulated microcalcifications: a receiver operating characteristic study based on phantom images.

RATIONALE AND OBJECTIVES The authors compare a 43-micron computed radiographic system with a mammographic screen-film system for detection of simulated microcalcifications in an observer-performance study. MATERIALS AND METHODS The task of detecting microcalcifications was simulated by imaging aluminum wire segments (200-500 microns in length; 100, 125, or 150 microns in diameter) that overlapped with tissue background structures produced by beef brisket. A total of 288 such simulations were generated and examined with both computed radiography and conventional screen-film mammography techniques. Computed radiography was performed with high-resolution plates, a 43-micron image reader, and a 43-micron laser film printer. Computed radiographic images were printed with simple contrast enhancement and compared with screen-film images in a receiver operating characteristic study in which experienced readers detected and scored the simulated microcalcifications. Observer performance was quantitated and compared by computing the area under the receiver operating characteristic curve. RESULTS Although the resolution of the computed radiography system was better than that of commercial systems, it fell short of that of screen-film systems. For the 100-micron microcalcifications, the difference in the average area under the curve was not statistically significant, but it was significant for the larger simulated microcalcifications: the average area under the curve was 0.58 for computed radiography versus 0.76 for screen-film imaging for the 125-micron microcalcifications and 0.83 versus 1.00, respectively, for the 150-micron microcalcifications. CONCLUSION Observer performance in the detection of small simulated microcalcifications (100-150 microns in diameter) is better with screen-film images than with high-resolution computed radiographic images.

EFFECTIVENESS OF ANTISCATTER GRIDS IN DIGITAL RADIOGRAPHY : A PHANTOM STUDY

INTRODUCTION.With ever-increasing interest in bedside radiography using digital imaging techniques, one question often asked is whether and which antiscatter grid should be used to reduce the effects of scattered radiation. In this article, the authors quantitatively analyze and compare images obtained with a flexible circular hole (FCH) grid, a conventional 6:1 focused grid, or without a grid. METHODS.Scatter-to-primary ratios (SPRs), contrast signals, and contrast-to-noise ratios (CNRs) were measured for fixed patient exposure and were compared at various locations in the images of an anthropomorphic chest phantom. RESULTS.Although both grids resulted in greater contrast signals, a significant improvement of the CNR was achieved only in the upper and middle mediastinum regions when the conventional grid was used. With the FCH grid, the CNR is degraded at most locations in the chest image. DISCUSSION.Our results indicate that at fixed exposure level, the use of either grid yields only marginal improvement, at best, in bedside digital chest radiography.

Signal fading, erasure, and rescan in storage phosphor imaging

This paper addresses three problems in storage phosphor imaging: natural fading of latent image signals, proper erasure of the exposed plates, and re-scanning for a second readout. Signal, and signal-to-noise ratios were measured as a function of time, erasure power/time, or number of pre-scans to study these problems. The latent image signals were found to decay very rapidly during the first several minutes and stabilize after several hours. The fading effect results in a variable signal gain (signal per unit exposure) which may affect the system calibration and quantitative use of the image data. Complete erasure of the latent image signals is necessary to ensure that no residual image signals are present when the plate is exposed again. It was found that plates used in high exposure applications (GI, therapeutic imaging) may require an excessively long erasure time to prepare them for use in low exposure applications (chest imaging). Although the latent image is partially erased during the readout process, it may sometimes become necessary to re-scan the plate for a second or third readout. It was found that because a large number of energy traps are generated for each x-ray photon, a significant portion of the x-ray information remains intact for reuse after the first or second scans. Measurement of the signals and signal-to-noise ratios are presented to demonstrate and discuss the aforementioned problems or effects.

Frequency-dependent DQE in dual-screen CR imaging

It has been shown that dual-screen image acquisition technique can be used to improve the image signal-to-noise ratio (SNR) in computed radiography (CR) imaging. In chest imaging situations, acquisition with a high resolution (HR) screen and a standard resolution (ST) screen can also be used to improve the modulation transfer function. Unlike in conventional radiography using two screens, the front and back images in dual-screen CR imaging can be separately read out and superimposed with the weighting factors selected to optimize a specific image quality descriptor. The purpose of this paper is to determine the weighting method which would optimize the frequency dependent detective quantum efficiency (DQE) in dual-screen CR imaging with an HR and a ST screen. A theoretical model is derived to relate the DQE in the superimposed image to those in the front and back images and to determine the optimal weighting factors and the maximum DQE that can be achieved. Using this model and DQEs measured for the HR and ST screens, we could estimate optimal weighting factors and maximum DQEs as a function of frequency. Various screen combinations were studied and compared for the maximum DQE that can be achieved. We have shown that for maximum DQE, the front and back images should be weighted in such a way that their magnitudes are proportional to the DQE divided by the MTF. The maximum DQE in the optimally superimposed image is equal to the sum of the DQEs of the front and back images.

Pixel averaging versus digitization using larger apertures: a comparison of the spatial resolution properties

2K (2048 X 2500) or 1K (1024 X 1250) digitized chest film images can be generated by either direct digitization or converting a 4K (4096 X 5000) digitized film image by pixel averaging. In this paper, these two methods are compared for their implication on the resolution properties of the resulting images. A film image of the lead bar resolution pattern was used as the source of all digital images. Signal profiles of the bar pattern were studied to compare the pixel averaging and direct digitization methods. Based on this comparison, it was found that pixel averaging, when used with proper filtering, can be used to simulate direct digitization using larger (210/420) apertures and result in similar square wave response. Pixel averaging, however, can result in better square wave response when 2 to 1 or 4 to 1 straight averaging is used or a sharper kernel is used in pre-filtering.

Comparison of three different schemes for dual-energy subtraction imaging in digital radiography: a signal-to-noise analysis

Dual-energy subtraction imaging techniques allow the tissue and bone structures in the patient to be imaged separately, thus removing some obscurity resulted from the overlapping of the two structures. Furthermore, they provide the potential for the tissue or bone contents to be quantified for diagnostic use. Thus, capabilities for dual-energy subtraction imaging are often incorporated with new digital radiography techniques. There are three different schemes for implementing dual-energy subtraction imaging techniques. Among them, dual-kVp and sandwich detector approaches are two most often used schemes. A third scheme is the single kVp-dual filter approach which allows a more flexible control of the spectra while avoiding kVp switching. It is suitable for digital radiography techniques using two linear detector arrays. In this paper, the signal-to-noise properties of these three schemes is computed for various combinations of kVp, filters and patient thicknesses (tissue and bone). Based on the signal-to-noise analysis, they are compared to each other for the efficiency of x ray usage, dose efficiency, and accuracy for background subtraction and thickness measurement.

Digital x-ray imaging with two-dimensional charge-coupled device arrays

We are investigating the characteristics of a prototype digital radiography imaging system in which six two-dimensional diode arrays (CCD) are directly coupled through a bonded matrix (3 x 2) of fiberoptic minifiers to either a scintillating fiberoptic glass plate or to a fluorescent screen. Images are digitally acquired at a rate of up to 30 frames/sec and software corrected for pixel gain, dark current, and geometric distortion. This paper describes the concepts and design configuration of this approach, as well as preliminary results from several phantom and animal studies. Our results indicate that high resolution (> 4 lp/rnm) and high signal-to-noise ratio images can be obtained with this method. However, the complexity associated with this concept cannot be discounted.

Regional SNR optimization in dual-screen computed radiography (CR) imaging

Dual-screen CR imaging technique can be used to improve x- ray absorption and image signal-to-noise ratio (SNR). Previous research has shown that it is possible to optimize the image superimposition process for best SNR in the superimposed image. Optimal weighting factors and SNRs have ben theoretically derived and related to the SNRs in the front and back images. The relationship has been experimentally verified. Practical implementation of the technique, however, involves non-uniform image signals and SNRs. To understand how the optimal weighting factors can be determined and how they vary spatially in non-uniform images, theoretical, numerical and experimental studies have been conducted. The results indicate that spatial variation of optimal weighting factors is minimal. An empirical method can be used to determine the optimal weighting factors over regions with minimal signal non-uniformity. The resulting weighting factors can be applied to superimposition of the entire front and back images for best SNR.