The development of the highest resolution, large-area, active- matrix, flat-panel, imager (AMFPI) thus far reported is described. This imager is based on a 97 micrometer pixel pitch array with each pixel comprising a single a-Si:H TFT coupled to a discrete a-Si:H n-i-p photodiode. While the initial configuration chosen for fabrication is a 2048 X 2048 pixel array, a larger monolithic array format of 3072 X 4096 pixels is also permitted by the design. When coupled to an overlying scintillator such as a phosphor screen or CsI:Tl, the array allows indirect detection of incident radiation. The array is operated in conjunction with a recently completed electronic acquisition system featuring asynchronous operation, a large addressing range, fast analog signal extraction and digitization, and 16-bit digitization. This imager, whose empirical characterization will be reported in a subsequent paper, was developed as an engineering prototype to allow investigation of the performance limits of the most aggressive array designs permitted by present active-matrix technology. The development of this new imager builds upon knowledge acquired from the iterative design, fabrication, and quantitative evaluation of earlier engineering prototypes based on a series of 127 micrometer pitch arrays. This paper summarizes the general program of research leading to this new device and puts this in the context of world-wide developments in indirect and direct detection AMFPI technology. Some limitations of present AMFPI technology are described, and possible solutions are discussed. Specifically, the incorporation of multiplexers based on poly-crystalline silicon circuitry into the array design, to facilitate very high resolution imagers, are proposed. In addition, strategies to significantly improve AMFPI performance at very low exposures, such as those commonly encountered in fluoroscopy, involving the reduction of additive noise (such as through lower preamplifier noise) and the enhancement of system gain (such as through the use of lead iodide) are discussed and initial calculations illustrating potential levels of performance are presented.

Progress is discussed on the improvement of a Direct Radiography^TM solid state, flat panel, digital detector designed for use in general radiographic applications. This detector, now known as DirectRay^TM, operates on the principle of direct detection of X-ray photons with a selenium photoconductor and consists of 500 micrometer thick amorphous selenium coupled to an amorphous silicon thin-film-transistor (TFT) readout array. This device is fabricated with a 14 X 17-inch (35 X 43-cm) active imaging area, corresponding to 2560 X 3072 pixels having dimensions of 139 micrometer X 139 micrometer and a geometrical fill factor of 86%. Improvements include a TFT array design upgrade with reduced noise characteristic, lower-noise readout electronics, and improved interfaces. Clinical radiographic images are currently being generated with the DirectRay detector using an X-ray exposure level equivalent to that of a 400 speed screen- film combination while maintaining the superior spatial resolution that is inherent in the direct conversion method. An effective sensor restoration technique has been implemented that eliminates the potential for selenium memory artifacts after a high dose. New results on NPS, MTF, DQE and signal linearity are presented. Detectability of low contrast objects using FAXiL test objects as well as the results of clinical studies are discussed.

We report the fabrication and evaluation of a Pbl₂ imager using large area amorphous silicon technology. This approach uses a thick Pbl₂ x-ray photoconductor to absorb x-rays and collect ionization charge under the action of an applied field, while amorphous silicon thin film transistors (TFT) provide a matrix-addressed read out of the signal to external electronics. The x-ray sensitivity of Pbl₂ is high, and mobility-lifetime product is large enough to yield a high charge collection at low applied fields. The test arrays used to evaluate Pbl₂ have 256 X 256 pixels of size 200 microns. Each pixel contains an amorphous silicon switching transistor, gate and data addressing lines, a charge storage capacitor and a metal pad to contact the Pbl₂ layer. Early evaluation of the image sensor indicates the promise of Pbl₂ but indicates that reduction of the leakage current is important.

A clinical imaging system based upon an amorphous-Silicon (a- Si) flat dynamic (digital) X-ray image detector (FDXD) has been developed. The objectives of this experimental set-up were to determine the physical image quality and to establish the clinical feasibility of a flat-panel x-ray detector for radiography and fluoroscopy (R&F) applications. The FDXD acquires dynamic X-ray images at high frame rates in both continuous and pulsed fluoroscopic modes, lower frame rate exposures and single shots. The system has been installed in a clinical research room at The General Infirmary, Leeds (UK) and is being evaluated in a variety of universal R&F contrast medium aided examinations, including barium swallows, meals and enema examinations. In addition, general radiographic examinations have been performed. Both the established benefits and possible drawbacks of this type of system, together with the potential solutions, are discussed in this paper. Approach, design and set-up of the system are presented, and the dose efficiency and image quality achieved in clinical operation are explained. The technical and medical phantom images have been evaluated and analyzed. The results of the clinical examinations in mixed applications are discussed. The results of the measurements and examinations performed to date on this experimental FDXD system confirm the potential of this new type of digital X-ray image detector.

A new 17' X 17' immediate direct digital flat panel detector has been developed to fit the needs of General Radiography. After reviewing a few key aspects of the General Radiography needs (X-ray energy range and associated measurement conditions, system integration and system operation), we describe the new detector Cesium Iodide/Amorphous Silicon based technology, and give measurement results (MTF, DQE stability). We compare the new detector performance to existing technologies (film/screen combination, storage phosphor devices) and also to other flat panel solutions (Selenium). We conclude that the CsI/a-Si technology is now the best suited one in order to fit the needs of General Radiography, this means all kinds of examinations (chest, abdomen, bones, extremities. . .) which have been up to now done using films.

The use of a diffraction spectrometer developed by Deslattes for determination of mammographic kV is extended to the measurement of accurate, relative x-ray spectra. A series of x-ray spectra were acquired by passing an x-ray beam through a bent-silicon diffraction crystal, and the diffracted x-rays were detected by a charge coupled device (CCD) coupled to an x-ray scintillating screen. The raw spectra were corrected both on the energy axis and on the x-ray photon fluence axis. The measured, corrected x-ray spectra were compared against tabulated x-ray spectra measured under similar conditions. Results indicate that the current device is capable of producing accurate relative x-ray spectra in the mammography energy range.

Quasi-monochromatic X-rays with energy in the 16 - 24 keV range have been produced in our laboratory by making use of a conventional W-anode X-ray tube and a monochromator optical system via Bragg diffraction. The optical system is based on an array of mosaic crystals which produces an irradiation field obtained with adjacent reflected beams. The field size is about 10.5 X 12.0 cm² in the image plane. The introduction of an active optical component such as a mosaic crystal array may modify the imaging performances of the system. In order to investigate how the resolution properties of the quasi-monochromatic source are affected by the use of adjacent beams and the Bragg diffraction phenomenon, images of a hole pattern and a slit camera have been obtained with different magnification factors. The images of the hole pattern show duplication effects due to the partial superposition of adjacent beams. Slit camera images show an unexpected increase of the focal spot blurring along the diffraction plane. This behavior could be explained by considering each crystal monochromator of the array as a secondary source and consequently as a focal spot of the beam coming from it. Along the orthogonal direction the image blurring depends on the focal spot size of the X-ray tube.

The constructions of a plasma flash x-ray generator having a cold-cathode radiation tube and its application to soft radiography are described. The x-ray generator employs a high- voltage power supply, a low-impedance coaxial transmission line with a gap switch, a high-voltage condenser with a capacity of 0.2 (mu) F, a turbo-molecular pump, a thyristor pulser as a trigger device, and a flash x-ray tube. The high- voltage main condenser is charged up to 60 kV by the power supply, and the electric charges in the condenser are discharged to the tube after triggering the cathode electrode. The flash x-rays are then produced. The x-ray tube is a demountable triode which is connected to the turbo molecular pump with a pressure of approximately 1 mPa. This tube consists of a rod-shaped carbon cathode, a trigger electrode made from a copper wire, a stainless-steel vacuum chamber, insulators, a polyethylene terephthalate x-ray window, and two anode electrodes (targets) of molybdenum and silver. The space between the anode and cathode electrodes had a constant value of approximately 20 mm, and the trigger electrode is set in the center of the cathode electrode. As the electron flows from the cathode electrode are roughly converged to the target by the electric field in the tube, the plasma x-ray source which consists of metal ions and electrons is produced by the target evaporating. Because the bremsstrahlung spectra are absorbed by the monochromatic filter, K-series characteristic x-rays are obtained. Both the tube voltage and current displayed damped oscillations, and their peak values increased according to increases in the charging voltage. In the present work, the peak tube voltage was almost equivalent to the initial charging voltage of the main condenser, and the peak current had a value of about 25 kA with a charging voltage of 60 kV. When the charging voltage was increased, the intensities of the K-series characteristic x-rays increased. Next, the intensities decreased as the monochromatic filter was inserted. Using this quasi-monochromatic plasma flash x- ray generator, we performed high-speed soft radiography with x-ray durations of about 1 microsecond.

X-ray spectra suitable for mammography, created by laser-based x-ray source at laser intensity about 10¹⁸ W/cm², where investigated. The spectra consisted of a continuous bremsstrahlung emission and discrete K_(alpha ), K_(beta ) lines and have been obtained for Mo, Rh, Ag, In and Sn targets (Z equals 42, 45, 47, 49, and 50) with K_(alpha ) emissions at 17.4, 20.2, 22.2, 24.2, and 25.3 keV, respectively. The continuous bremsstrahlung component extended to high energies with no cut-off energy. The shape of the continuous bremsstrahlung spectrum was described by the function E^-p, where p (greater than 0) was primarily determined by the temperature of hot electrons produced in laser beam- target interaction. The absolute value of x-ray yield was proportional to the atomic number of the target Z. The intensity of characteristic x-rays was about a factor of five higher than the corresponding bremsstrahlung, and showed weak dependence (increase) when the atomic number of the target increased from Z equals 42 to Z equals 50.

We are developing a slit-scanning digital mammography system with the goal of reducing radiation dose and improving dynamic range in comparison to conventional film-screen mammography. The system is based on a linear silicon detector connected to a fast parallel processing ASIC controlled from a PC. The detector system is configured from a silicon wafer such that the X-rays enter the wafer edge-on, resulting in a high efficiency over the entire range of diagnostic X-ray energies since the absorption length of the detector can be made large. An evaluation of the system gives an MTF close to that predicted theoretically for a 100-micrometer pitch pixel detector. We also propose to reduce tube-loading in the slit- scanning system with a refractive X-ray lens that can increase the incident photon flux by about 5-fold according to ray tracing simulations. Moreover the lens shapes the energy spectrum to obtain pseudo-monochromatic beam thus enhancing contrast sensitivity, for example in mammography. We estimate a scan time with the final system of the order of a few seconds for a full-field digital mammogram. Phantom images of microcalcifications and of tumor-like masses, and an initial evaluation of a scanned-slit photon-counting X-ray imaging system, are presented.

A new prototype imaging system based on ultrasound transmission through the object of interest -- acoustical holography -- was developed which incorporates significant improvements in acoustical and optical design. This system is being evaluated for potential clinical application in the musculoskeletal system, interventional radiology, pediatrics, monitoring of tumor ablation, vascular imaging and breast imaging. System limiting resolution was estimated using a line-pair target with decreasing line thickness and equal separation. For a swept frequency beam from 2.6 - 3.0 MHz, the minimum resolution was 0.5 lp/mm. Apatite crystals were suspended in castor oil to approximate breast microcalcifications. Crystals from 0.425 - 1.18 mm in diameter were well resolved in the acoustic zoom mode. Needle visibility was examined with both a 14-gauge biopsy needle and a 0.6 mm needle. The needle tip was clearly visible throughout the dynamic imaging sequence as it was slowly inserted into a RMI tissue-equivalent breast biopsy phantom. A selection of human images was acquired in several volunteers: a 25 year-old female volunteer with normal breast tissue, a lateral view of the elbow joint showing muscle fascia and tendon insertions, and the superficial vessels in the forearm. Real-time video images of these studies will be presented. In all of these studies, conventional sonography was used for comparison. These preliminary investigations with the new prototype acoustical holography system showed favorable results in comparison to state-of-the-art pulse-echo ultrasound and demonstrate it to be suitable for further clinical study. The new patient interfaces will facilitate orthopedic soft tissue evaluation, study of superficial vascular structures and potentially breast imaging.

Prostate cancer is the most commonly diagnosed cancer in men in North America. Transrectal ultrasound imaging (TRUS) is widely used for the evaluation of prostate disease. However, with conventional TRUS, diagnosticians must mentally integrate a series of two-dimensional (2D) images in order to develop an impression of the 3D anatomy, and the accurate estimation of prostate volume is difficult. We propose using 3D TRUS to overcome these problems. In this paper, we describe a 3D ultrasound imaging system and study its performance. The system consists of a conventional ultrasound machine, a microcomputer with a video frame grabber, and a custom-built assembly for rotating the probe. A typical scan of 200 2D B- mode images takes 13 seconds. These images can then be reconstructed into a 3D image, and the resulting 3D image can be interactively displayed using 3D visualization software. We also show that manual planimetry of 3D TRUS images can be used to estimate prostate volumes in vitro with 2.6% accuracy and 2.5% precision; and in vivo with 5.1% intra-observer variability and 11.4% inter-observer variability. Thus, 3D TRUS overcomes the limitations of 2D TRUS, and has great potential as a tool for the diagnosis and management of prostate disease.

The development of an exact and practicable technique that allows the full time-space complexity of a wideband ultrasound field to be imaged, from a single set of local measurements, is presented. The method is an improvement over the angular spectrum technique, provides a novel approach towards eliminating evanescent waves, and also allows for more efficient computation of transient fields. The new representation of a field has simply propagating components that may be directly measured (via a specifically designed hydrophone) as the time-varying spatial projections of the field. Reconstruction of the transient field is performed by a technique that is a generalization of the Fourier slice theorem. The theory is vindicated by explicit demonstrations with measurements of the fields from typical ultrasound transducers. Visualization of the field is either as a three- dimensional pressure distribution at any temporal instant, or as the time-variation of the pressure over any plane orthogonal to the field propagation direction. This new method for ultrasound field measurement, prediction, and visualization is shown to be eminently practicable and represents a substantial improvement over conventional methods.

Local motion of object in MR imaging usually leads to a motion artifact which is spread in phase coding direction. Subsequently, the spreading artifact results in the degradation of signal intensities even though static parts. A ghost artifact and a zipper-like artifact are examples for the object motion and flow effect. In this paper, a new local motion suppression technique is proposed where signal affected by the motion is localized using a short time Fourier transform or a windowed Fourier transform. Further, an MR pulse sequence for the windowed Fourier transform was developed and applied to object having local motions thereby showing local motion suppression. The computer simulations of the proposed method and their corresponding experimental results are reported.

Reconstruction aspects of spiral-scan magnetic resonance imaging are investigated with polar and rectangular coordinates based reconstruction. Several reconstruction algorithms based on these coordinates are tested, and reconstructed image qualities are compared with a mathematically built phantom. An improved reconstruction algorithm with dc-offset correction in projection domain is proposed, which provides the best reconstructed image quality from simulation. Image artifact with existing algorithms is also completely removed with the proposed method.

One of the goals of medical imaging scientists and bioengineers remains the development of digital electronic technologies that can replace film-based methods of acquiring x-ray images. With the achievement of this goal, all diagnostic imaging technologies would be based on digital techniques with all the attending benefits. Based on the performance of numerous research prototype small-field and one large-field Kinestatic Charge Detector (KCD) system for digital radiography, a large-field clinical KCD scanner is currently being designed and built for technical evaluation and for clinical evaluation of 200 volunteer patients (including clinical comparisons with film, storage phosphor, and other available clinical systems). The state of development of this clinical KCD system, including detector, data-acquisition system and scanning gantry design, is reviewed in this paper.

This paper provides a comparison of some imaging parameters of four portal imaging systems at 6 MV: a flat panel detector, two CCD cameras and an electron beam tube based video camera. Measurements were made of signal and noise and consequently of signal-to-noise per pixel as a function of the exposure. All systems have a linear response with respect to exposure, and with the exception of the electron beam tube based video camera, the noise is proportional to the square-root of the exposure, indicating photon-noise limitation. The flat-panel detector has a signal-to-noise ratio, which is higher than that observed with both CCD-Cameras or with the electron beam tube based video camera. This is expected because most portal imaging systems using optical coupling with a lens exhibit severe quantum-sinks. The measurements of signal-and noise were complemented by images of a Las Vegas-type aluminum contrast detail phantom, located at the ISO-Center. These images were generated at an exposure of 1 MU. The flat-panel detector permits detection of Aluminum holes of 1.2 mm diameter and 1.6 mm depth, indicating the best signal-to-noise ratio. The CCD-cameras rank second and third in signal-to- noise ratio, permitting detection of Aluminum-holes of 1.2 mm diameter and 2.2 mm depth (CCD_1) and of 1.2 mm diameter and 3.2 mm depth (CCD_2) respectively, while the electron beam tube based video camera permits detection of only a hole of 1.2 mm diameter and 4.6 mm depth. Rank Order Filtering was applied to the raw images from the CCD-based systems in order to remove the direct hits. These are camera responses to scattered x-ray photons which interact directly with the CCD of the CCD-Camera and generate 'Salt and Pepper type noise,' which interferes severely with attempts to determine accurate estimates of the image noise. The paper also presents data on the metal-phosphor's photon gain (the number of light-photons per interacting x-ray photon).

In a previous paper, a portal imaging system was described that used a 101 mm diameter, 3 mm thick CsI (Tl) transparent scintillating screen coupled to a liquid-nitrogen-cooled slow- scan CCD-TV camera with a 40 mm f1.0 macro lens with a 5:1 demagnification. Meanwhile, improved images have been acquired using a 50 mm f1.1 macro lens with a 7:1 demagnification. These images were presented at an AAPM International Symposium on Electronic Portal Imaging in Detroit, MI, in May, 1997. Since the Detroit meeting, a 203 mm diameter, 13 mm thick CsI(Tl) crystal has been purchased from Bicron. This transparent screen has been used with a Nikkor 35 mm f1.4 lens to show the whole 203 mm circular field at 0.53 mm pixel size with the existing Astromed liquid nitrogen cooled CCD TV camera system. The geometry of the imaging system has been optimized to achieve high spatial resolution (1 lp/mm) in spite of the increased thickness of the screen. This increased thickness allows the high image quality achieved with the older screen at 72 MU to be maintained with the newer screen while reducing the dose to 1 MU. Images have been acquired with the new screen of lead bar patterns, low-contrast hole patterns in Lucite blocks, and anthropomorphic phantoms.

A number of new concepts are presently appearing for large size X-ray imaging. The almost perfect MTF obtained with photoconductors read by an a-Si matrix is undoubtedly the best as regards signal, but the associated aliasing effects force the total photon noise power into the useful spatial frequency range, which degrades the signal to noise ratio. At present, an imager consisting of a CsI:Tl scintillator layer coupled to an array of a-Si pixels is the most attractive concept because it combines (1) A strong X-ray absorption together with a good MTF, (2) The best overall X-ray to electrons conversion factor, necessary to overcome electronic noises, (3) A low pass filtering effect which strongly attenuates photon noise beyond the Nyquist limit. The variation of DQE with spatial frequency is the right factor of merit for an X-ray imager. It not only accounts for the spatial resolution, but also contains the effects of all noises in the detection process, including aliasing. Whenever a digital imager is compared with an analogue detector such as the screen/film, the effect of the position of the object with respect to the pixel lattice (or phase difference between the sampling pitch a and the spatial frequency of interest f) must be taken into account. This results in an additional sinc²(fa) factor applied to the DQE(f).

Electronic portal imaging devices (EPIDs) have many potential advantages over portal films used in radiation therapy. The most commonly used EPIDs are video-based systems. Camera noise is one of the main noise sources which limit the performance of current video-based EPIDs. In this paper, we investigate the use of an avalanche-multiplication-based video camera to effectively eliminate the camera noise in an EPID and, therefore, to improve the quality of megavoltage portal images. To make these investigations, a new EPID with the avalanche-multiplication-based video camera has been made. Detective quantum efficiency (DQE) of the new device has been measured with a ⁶⁰Co radiation beam. It is shown that the camera noise is much smaller than the quantum noise in the new EPID system and the DQE value of the system is significantly increased because of the elimination of camera noise.

This paper describes a novel area detector for direct conversion and readout of the x-ray energy that eliminates multiple conversions and coupling stages which degrade performance. The pixel array and readout electronics are fabricated on the same piece of silicon. The detector consists of a uniform layer (approximately 300 micrometers) of amorphous selenium alloy vapor-deposited on an electronic readout array fabricated using conventional complementary metal oxide semiconductor (CMOS). The CMOS array features 66 micrometer pixels in a 1024 X 832 array providing a 5.5 X 6.75 cm image area. Each pixel has active circuitry including signal amplification, pixel selection and reset, while peripheral circuitry on one end of the array provides shift registers, sample and hold and multiplexing. The CMOS readout array was fabricated at a standard facility on a 10-cm diameter silicon wafer using 2 micrometer CMOS process. Fifteen separate image sensors were assembled for evaluation in a 3 X 5 format to provide a 20 X 27 cm composite field of view. Missing data between sensors is recovered by acquiring three sub-exposures, between which the array is translated diagonally approximately 2 mm. Total exposure time for an average breast is less than one second. Conversion efficiency was found to be approximately 120 electrons per absorbed x-ray (19 keV average). Electronic readout noise was measured to be 2.4 ADU corresponding to approximately 500 electrons. Detective quantum efficiency was found to be 0.65 at low spatial frequency (0.25 lp/mm) and at 0.2 at high spatial frequency (8 lp/mm) for x-ray fluence ranging from 5 - 35 mR. Images of an ACR phantom show visualization of all of the fibers, specks and masses when displayed with a linear lookup table on a high-resolution monitor. These studies demonstrated that there is a slight but measurable image retention evident as 'ghost' images. The two most effective means to reduce this effect are flushing the sensors with infrared light or x-rays between exposures and reversing the applied voltage on the selenium layer. A number of improvements designed to increase sensitivity and reduce noise also have been identified and are being implemented. Sample images were acquired from four volunteer human subjects at exposure factors identical to their film-screen mammograms. The results suggest that the detector performance is suitable for further clinical investigation.

Current clinical mammographic imaging has been limited to screen/film, but advances in photodetector and x-ray converter technology has opened up the possibility of significantly improving mammographic screening with digital imagers. One limitation of some recently developed in direct detection imagers is a limited optical fill factor in the photodetector. A small fill factor reduces the amount of light collected from individual x-ray interactions in the screen. It can also lead to an overall reduction in the number of x-ray quanta detected. In addition, the fill factor usually decreases with an increase in photodetector resolution. To reduce these effects, we propose a technique for improving the effective fill factor of any indirect detection imager by incorporating focusing microlens arrays between the phosphor screen and the photodetector. In this investigation, we evaluate the light collection efficiencies of our proposed imager/microlens combination using Monte Carlo simulation and optical ray tracing.

Linear-systems theory can be used to characterize the performance of many imaging systems in terms of signal- transfer and noise-transfer relationships. Using this approach, complex systems are described as serial cascades of simple processes. In a series of articles by Shaw, Rabbani, Van Metter, Barrett, Wagner and others, key processes have been identified and relationships developed which describe the transfer of the auto-covariance function and noise-power spectrum (NPS). However, to date only serial cascades have been described. In this article, this approach is extended to also include parallel cascades under certain conditions. Parallel cascades are used to describe systems in which the output signal is the sum of two or more serial cascades. The output NPS is therefore the sum of the NPS from each serial cascade plus cross-spectral density terms which are required to account for statistical correlations between the serial cascades. An expression for the cross-spectral density term is developed for the special case of a serial cascade branching into two parallel cascades at a point where image quanta are uncorrelated. This work was inspired by an article published by Metz and Vyborny who showed the effect of reabsorption of characteristic x rays on the NPS of radiographic screens using a complex statistical analysis. The linear-systems approach is used here to derive the same result making use of a 'flow diagram' which represents the sequence of events giving rise to light emission in a radiographic screen as three serial cascades of stochastic point processes. For many, the flow diagram approach is more readily understood than a statistical analysis, and may offer more physical insight into an understanding of the results.

In the clinical setting, image quality is most commonly evaluated by the visual observation of images of test objects and/or phantoms. Because of the uncertainties in such results (either large variance or bias or both), more precise quantitative measures based on statistical decision theory should be investigated. A series of simulations and experiments were conducted to investigate the statistical properties, i.e., the bias and variance, of the estimate of the square of the SNR of the 'ideal' observer (SNR_PWMF). Several methods of bias reduction were compared including one due to Fukunaga and Hayes. Good agreement was obtained between the results of simulations and the theoretical predictions for the bias and variance. The different methods of bias reduction have the same applicability for both 'ideal' and 'quasi-ideal' observers for the series of SKE/BKE tasks investigated in the present study. This work also provides some new avenues for additional investigation. First, the techniques can lead to protocols for making the evaluation of imaging system performance with a limited number of sample images, which is an important issue for any clinical implementation. Second, since selective spatial frequency channels can be used in estimating the SNR_PWMF, the method has potential utility for imaging tasks beyond SKE/BKE tasks such as those with clinically relevant backgrounds but possessing stationary statistics.

A technique to estimate the radial dependence of the noise power spectrum of images is proposed in which the calculations are conducted solely in the spatial domain of the noise image. The noise power spectrum averaged over a radial spatial- frequency interval is obtained from the variance of a noise image that has been convolved with a small kernel that approximates a Laplacian operator. Recursive consolidation of the image by factors of two in each dimension yields estimates of the noise power spectrum over that full range of spatial frequencies.

To produce a parallel-projection image on a planar array of gamma-ray detectors, the standard procedure is to use a parallel-hole collimator. An alternative presented here to place a pinhole array in front of the detector and to synthesize parallel projections from the pinhole data. This eliminates the physical blurring and distance-dependent resolution effects of collimators, and replaces them with the mathematical problem of estimating projection integrals. We provide conditions under which this can reasonably be done and show some examples.

A method for analyzing the effects of quantization, developed for temporal one-dimensional signals, is extended to two- dimensional radiographic images. By calculating the probability density function for the second order statistics (the differences between nearest neighbor pixels) and utilizing its Fourier transform (the characteristic function), the effect of quantization on image statistics can be studied by the use of standard communication theory. The approach is demonstrated by characterizing the noise properties of a storage phosphor computed radiography system and the image statistics of a simple radiographic object (cylinder) and by comparing the model to experimental measurements. The role of quantization noise and the onset of contouring in image degradation are explained.

A method is developed which makes it possible to scan and reconstruct an object with cone beam x-ray in a spiral scan path with area detectors much shorter than the length of the object. The method is mathematically exact. If only a region- of-interest of the object is to be imaged, a top circle scan at the top level of the region-of-interest and a bottom circle scan at the bottom level of the region-of-interest are added. The height of the detector is required to cover only the distance between adjacent turns in the spiral projected at the detector. To reconstruct the object, the Radon transform for each plane intersecting the object is computed from the totality of the cone beam data. This is achieved by suitably combining the cone beam data taken at different source positions on the scan path; the angular range of the cone beam data required at each source position can be determined easily with a mask which is the spiral scan path projected on the detector from the current source position. The spiral scan algorithm has been successfully validated with simulated cone beam data.

CT 2(pi) helical weighting algorithms do not lend themselves to fast reconstruction: the weight distributions present a line of discontinuity across the sinogram which defines two separate regions and associated weight expressions. Accordingly, reconstruction of P image planes requires P weightings and filterings of all projections. This paper shows that, by generalizing the concept of the interpolation/extrapolation function to that of distance function, and by selecting particular classes of such functions, the sinogram discontinuity can be eliminated. By imposing specific necessary conditions, single analytical expressions across the entire 2(pi) sinogram are obtained. Decomposition of these particular 'single' functions leads to exact or approximate fast two-filtering algorithms, for which a given projection needs to be filtered only two times for an arbitrary number P of reconstruction planes. Further, another generalization of the concept of helical weighting leads to 'single' weighting functions that depend only on the sum of the projection- and fan-angles. Accordingly, after rebinning the fan-beam projections to parallel projections, weighting commutes with filtering, and reconstruction of an arbitrary number P of image planes requires only one filtering per projection.

Interventional procedures with the guidance of computed tomography have attracted significant interest among clinicians in recent years. The application of computed tomography fluoroscopy (CTF) to interventional procedures has not only significantly improved the patient safety and reduced the probability of procedure complication, but also significantly shortened the procedure time. Recent studies, however, have indicated that a significant time delay exists in the CTF due to the nature of tomographic reconstruction process. This inherent limitation could lead to either confusions to the operator or prolonged interventional procedure time. A detailed comparison of various tomographic reconstruction algorithms indicates that the halfscan algorithm provides one of the best performance in terms of time delay. To further improve the temporal response of the CTF system, we propose an optimized halfscan (OHS) algorithm. The OHS algorithm suppresses information at the start of the scan by applying non-positive weights to the projections. At the same time, the projections near the end of the scan are given weights larger than unity. The weighting function is designed such that the overall contribution from the redundant samples remains constant. The parameters of the weighting function are selected to achieve a proper balance between the temporal response of the system and the acceptable level of image artifacts. Theoretical analysis and phantom experiments have shown that a significant improvement in CTF temporal response can be achieved with the proposed scheme.

Problems in visualizing the complex anatomy of the cerebral vasculature during intravascular embolization therapy remain due to the two-dimensional nature of digital subtration angiography. We describe the characterization of a three- dimensional (3-D) imaging technique, Computed Rotational Angiography (CRA). Projection images were acquired by rotating a modified Siemens Multistar Angiographic prototype system (C- arm mounted XRII) around the object, resulting in approximately 130 images over 200 degrees in less than 5 s. Exposure time is less than 20 ms/frame; tube voltage ranges from 73 - 110 kVp; tube current ranges from 100 - 500 mA. In vitro resolution was tested using both small area (line patterns) and large area (beads) phantoms. Investigations using in vivo porcine models examined SNR in the presence of physiological flow conditions. Limiting high-contrast resolution was better than 6.2 lp/cm. Reconstruction of the large area phantom demonstrated uniform image quality. A comparison of model and measured SNR showed good agreement for low dose but significant difference for high dose reconstructions. SNR was 60 in multi-planar reformatted slices, and 140 in the maximum intensity projection through the same volume. In conclusion, the CRA technique we describe, -- combined with recent advances in computing hardware -- make the presentation of 3-D volumes in under 5 minutes during interventional procedures a very real possibility.

A Selenium thin film transistor (STFT) array-based rotational volume tomographic digital angiography (VTDA) imaging system has been constructed. The system consists of an x-ray tube and a STFT detector that are separately mounted on a gantry. This system uses the STFT array as a two-dimensional (2D) detector so that a set of 2D projections can be acquired for a direct three-dimensional (3D) reconstruction. This paper presents the results of the preliminary phantom studies using the STFT- based volume tomographic angiography imaging system. This research work demonstrates through phantom studies that the STFT array introduces no distortion and the STFT-based VTDA imaging system has a higher spatial resolution and better contrast resolution than an image intensifier(II)-based system.

Blur and noise in radiographs are caused in part by the transport of fluorescent radiation in imaging detectors. We have studied this effect using a Monte Carlo radiation transport analysis which tracks radiations associated with K, L, and M shell transitions. Energy deposition distributions are accumulated which permit computation of the large area energy absorption and noise characteristics. Additionally, the spatial distribution of deposited energy is evaluated in a manner which permits determination of the line spread function and the auto-correlation function. The frequency dependent detective quantum efficiency, D_QE(E,f), is subsequently determined by Fourier analysis. This novel method is illustrated by considering the response of a selenium direct digital detector to 120 keV x-rays. It is shown that fluorescent radiations associated with the composition of glass substrates cause a frequency dependent drop in D_QE(E,f)/D_QE(E,O) of 10% to 22%.

This study was performed to investigate the effect of an additional antiscatter grid on the detail signal-to-noise ratio (SNR) and the visual detail detectability with a digital chest radiography system. An anthropomorphic chest phantom with different types of superimposed lesions was used to obtain four series of images simulating slim and thick patients, both with and without an additional grid. The exposure to the phantom was identical for the grid and non- grid situations. Difference images were then produced by subtracting an 'empty' thorax image (without superimposed lesions) from each image of the series. The difference images were used to measure detail contrast and detail SNR in different areas of the chest. Although the grid generally improved the lesion contrast, an improvement in lesion SNR was only obtained for some lesions in the obscured regions. In the lung area the lesion SNR was lower with the grid. ROC analysis showed only minor differences between grid and non-grid images in the visual detectability of the lesions; detectability was significantly higher, however, in the difference images compared to the original chest images. The results indicate that the use of the additional antiscatter grid is not necessary if the image contrast is restored by suitable image processing. Methods which reduce or eliminate the overlying anatomical structures in chest images lead to a significantly higher lesion detectability.

The effects of the stationary noise patterns and variable pixel responses that commonly occur with uniform exposure of digital detectors can be effectively reduced by simple 'flat- field' image processing methods. These methods are based upon a linear system response and the acquisition of an image (or images) acquired at a high exposure to create an inverse matrix of values that when applied to an uncorrected image, remove the effects of the stationary noise components. System performance is optimized when the correction image is totally free of statistical variations. However, the stationary noise patterns will not be effectively removed for flat-field images that are acquired at a relatively low exposure or for systems with non-linear response to incident exposure variations. A reduction in image quality occurs with the incomplete removal of the stationary noise patterns, resulting in a loss of detective quantum efficiency of the system. A more flexible approach to the global flat-field correction methodology is investigated using a pixel by pixel least squares fit to 'synthesize' a variable flat-field image based upon the pixel value (incident exposure) of the image to be corrected. All of the information is stored in two 'equivalent images' containing the slope and intercept parameters. The methodology provides an improvement in the detective quantum efficiency (DQE) due to the greater immunity of the stationary noise variation encoded in the slope/intercept parameters calculated on a pixel by pixel basis over a range of incident exposures. When the raw image contains a wide range of incident exposures (e.g., transmission through an object) the variable exposure flat-field correction methodology proposed here provides an improved match to the fixed-point noise superimposed in the uncorrected image, particularly for the higher spatial frequencies in the image as demonstrated by DQE(f) measurements. Successful application to clinical digital mammography biopsy images has been demonstrated, and benefit to other digital detectors appears likely.

This study investigated the signal to noise ratio (SNR) and radiation dose as a function of photon energy in screen-film mammography. An analytical expression was derived for the SNR for monoenergetic photons incident on simulated masses and microcalcifications embedded in uniform slabs of Lucite. The mean dose, D, was determined by dividing the total energy deposited in the Lucite phantom by the corresponding phantom mass. SNR and dose data for different photon energies were normalized to a constant value of energy absorbed by a 34 mg/cm² Gd₂O₂S screen. A figure of merit (FOM), defined as SNR²/D, permitted the optimum photon energy to be determined for each imaging task. For microcalcifications, the optimum energy was dependent on the size of the microcalcification, and increased from 19 keV for 100 micrometer to 22 keV for 500 micrometer imaged in a 4 cm Lucite phantom. The optimal photon energy for microcalcifications was 16 - 19 keV for a 2 cm phantom, increasing to 24 - 26 keV for an 8 cm phantom. For simulated masses of all diameters (2 mm to 10 mm) and thickness (0.15 to 0.6 mm), the optimal photon energy was approximately 17 keV for a 2 cm phantom, and increased to approximately 32 keV for an 8 cm phantom.

Coded aperture techniques based on a cyclic difference set uniformly redundant array (URA) can increase sensitivity of an imaging system without degrading the spatial resolution. In this paper, we discuss the pattern design and present its application for diagnostic nuclear medicine imaging with experimental results. Point-like, planar, and three- dimensional 140 keV gamma-ray sources are used in our experiments. We have experimentally demonstrated a three- dimensional coded aperture technique for nuclear medicine imaging and have compared it with conventional collimator systems.

This paper describes a multi-mode, digital imager for real- time x-ray applications. The imager has three modes of operation: low dose fluoroscopy, zoom fluoroscopy, and high resolution radiography. These modes trade-off resolution or field-of-view for frame rate and additionally optimize the sensitivity of the imager to match the x-ray dose used in each mode. This large area sensing technology has a form factor similar to that of a film cassette, no geometric image distortion, no sensitivity to magnetic fields, a very large dynamic range which eliminates repeat shots due to over or under exposure, 12 bit digital output and the ability to switch between operating modes in real-time. The imager, which consists of three modules: the Receptor, the Power Supply and the Command Processor, is intended as a component in a larger imaging system. Preliminary characterization of the prototype imager in fluoroscopic mode at entrance exposure rates down to 2.5 (mu) R/frame, indicates that the DQE(f), MTF and low contrast resolution are comparable to that obtained with an image intensifier tube (IIT) coupled to a video camera.

Flat-panel detector (FPD) is the driving force for realizing the next generation of x-ray systems. The purpose of this study was to develop a selenium-based FPD for both radiography and fluoroscopy. The detector uses amorphous selenium (a-Se) and a thin-film transistor (TFT) array. The simple construction of the a-Se layer permits real-time readout. The unique response characteristics of the FPD, which can be saturated over permitted x-ray doses, are provided by the TFT structure. Our prototype FPD was designed to acquire images at 30 frames per second (fps). A high modulation transfer factor was obtained: 0.63 at 2.0 Lp/mm. Sequential fluoroscopic images were acquired at up to 30 fps. The linear characteristics of the detector covered the commonly employed range of clinical exposure dose. Less than 1.5% image lag was measured at 30 fps.

In this paper we address design concepts and performance characterization with our laboratory x-ray detector system. Key component is a 1k² pixel TFT switched photodiode array with a pixel pitch of 200 micrometer. It is built of a-Si:H with a CsI:Tl scintillator layer. The detector system can be used for radiography and fluoroscopy applications with up to 30 images/s. It shows a S/N-ratio better than 23dB at a dose of 10nGy/frame and reaches a value for DQE of more than 60% at low spatial frequencies. We have developed a new evaporation process for CsI:Tl deposition directly on the array. It yields an x-ray sensitivity close to the theoretical limit and a spatial resolution on a sufficiently high level. An optimized plate design in combination with a dedicated charge sensitive readout amplifier chip lead to a very low level of electronic noise. In particular sources and properties of electronic noise and signal crosstalk have shown to be crucial for the clinical use of the new technology. The visual impression of the remaining noise in the images from our system is isotropic. This means especially that synchronous noise has been reduced to the edge of visibility.

Our goal is to develop a large area, flat panel solid-state detector for fluoroscopy. The detector employs a layer of photoconductor to convert incident x-rays directly to a charge image, which is then read out in real-time using a two dimensional array of thin film transistors (TFTs), or 'active matrix.' In order to guide the design of an optimum fluoroscopic flat-panel detector, a cascaded linear systems model was developed, from which the spatial frequency dependent detective quantum efficiency [DQE(f)] can be obtained. Then DQE(f) was calculated as a function of different detector design parameters, e.g. pixel fill-factor, x-ray exposure, Swank factor, electronic noise, and the calculation was performed for three different x-ray photoconductors: amorphous selenium (a-Se), cadmium zinc telluride (CZT), and lead iodide (PbI₂). A critical comparison was made of the advantages and disadvantages of each photoconductor. The results showed that the DQE(0) of all direct detectors has a linear dependence on the pixel fill- factor. For an a-Se layer with an electric field of 10 V/micrometer, DQE(f) is significantly degraded by the electronic noise of the detector, especially at very low x-ray exposure rates (e.g. 0.1 (mu) R/frame). With CZT and PbI₂, the detector is more tolerant of electronic noise because of the larger number of charge generated for each absorbed x-ray. We have applied our cascaded linear systems model of the direct, flat-panel detector to fluoroscopy. The theoretical predictions of DQE(f) for different detector parameters, e.g. the type of x-ray photoconductor, fill-factor, and electronic noise, provide a guideline for an optimum detector design for fluoroscopy.

In reference 1 we have presented the principle of an X-ray detector based upon a screen coupled to an array of multiple CCD sensors. In reference 2 we focus on the characterization of the image quality: resolution (MTF) and noise behavior in the overlap area. Simple (and cheap) low F# lenses likely show distortion which means that not all imaged pixels have the same magnification. This may affect resolution. Lenses with (some) barrel distortion have the benefit of less vignetting. The correction of distortion in combination with a rotation adjustment requires interpolation. Interpolation affects the noise properties so care must be taken in order to avoid that the noise characterization of the reconstructed image mosaic i.e. the noise texture becomes spatially non uniform. We present an analysis of the influence of lens distortion and interpolation in cases of small rotation correction on the image mosaic. The image processing appears not to diminish the image quality provided the processing parameters are set correctly. The calibration of the imaging mosaic geometry is crucial. We therefore present a robust extraction algorithm. In this paper our main interest is on MTF and quantum noise properties. The lab prototype hardware is designed such (cubic spline interpolation) that also the lens distortion can be compensated. For this purpose ASICs are designed by the company AEMICS. This enables relative cheap optical components with low F# and a short building length. We have obtained and will present radiographic exposures of static phantoms.

An evaluation of the physical imaging performance of a prototype CCD-based TV camera (XTV16) cardiac Digital Fluorography system is presented. A tube-based TV camera (XTV11) operates in parallel, via a 50% mirror, allowing a direct comparison between the two different TV image recording technologies. The MTF, Noise Power Density (NPD) spectrum and the DQE of the system have been determined. The NPD analysis has been completed in both horizontal and vertical directions and, for completeness, a two dimensional noise analysis of the system has also been carried out. An audit of the main sources of noise in the systems is presented. The effectiveness of image corrections in minimizing systematic noise due to the CCD camera is demonstrated. The DQE spectra of both systems at zero frequency are X-ray quantum noise limited and they are both operating dose efficiently. The DQE spectrum in the horizontal direction of the XTV16 at high spatial frequencies is shown to be superior to that of the XTV11 which may translate to improved rendition of small features in clinical images. The CCD camera system described here s now used in the Cardio-Vascular systems of a major European company.

This paper introduces new high-resolution amorphous Silicon (a-Si) image sensors specifically configured for demonstrating film-quality medical x-ray imaging capabilities. The devices utilizes an x-ray phosphor screen coupled to an array of a-Si photodiodes for detecting visible light, and a-Si thin-film transistors (TFTs) for connecting the photodiodes to external readout electronics. We have developed imagers based on a pixel size of 127 micrometer X 127 micrometer with an approximately page-size imaging area of 244 mm X 195 mm, and array size of 1,536 data lines by 1,920 gate lines, for a total of 2.95 million pixels. More recently, we have developed a much larger imager based on the same pixel pattern, which covers an area of approximately 406 mm X 293 mm, with 2,304 data lines by 3,200 gate lines, for a total of nearly 7.4 million pixels. This is very likely to be the largest image sensor array and highest pixel count detector fabricated on a single substrate. Both imagers connect to a standard PC and are capable of taking an image in a few seconds. Through design rule optimization we have achieved a light sensitive area of 57% and optimized quantum efficiency for x-ray phosphor output in the green part of the spectrum, yielding an average quantum efficiency between 500 and 600 nm of approximately 70%. At the same time, we have managed to reduce extraneous leakage currents on these devices to a few fA per pixel, which allows for very high dynamic range to be achieved. We have characterized leakage currents as a function of photodiode bias, time and temperature to demonstrate high stability over these large sized arrays. At the electronics level, we have adopted a new generation of low noise, charge- sensitive amplifiers coupled to 12-bit A/D converters. Considerable attention was given to reducing electronic noise in order to demonstrate a large dynamic range (over 4,000:1) for medical imaging applications. Through a combination of low data lines capacitance, readout amplifier design, optimized timing, and noise cancellation techniques, we achieve 1,000e to 2,000e of noise for the page size and large size arrays, respectively. This allows for true 12-bit performance and quantum limited images over a wide range of x-ray exposures. Various approaches to reducing line correlated noise have been implemented and will be discussed. Images documenting the improved performance will be presented. Avenues for improvement are under development, including higher resolution 97 micrometer pixel imagers, further improvements in detective quantum efficiency, and characterization of dynamic behavior.

We have developed a brand new, large-area X-ray image sensor for Digital Radiography System (DRS). The sensor utilizes a thin film transistor (TFT)/metal insulator semiconductor (MIS)-type photoelectric converter array made from hydrogenated amorphous silicon (a-Si:H). The sensor has 2688 X 2688 pixels at a pitch of 160 micrometer. The active area is 17 inch X 17 inch. The sensor utilizes scintillator coupled to the array. The light generated by X-rays is detected by the MIS-type photoelectric converters, and the resultant signals are scanned out by switching the TFTs. The a-Si TFT/MIS-type photoelectric converter array is characterized by high signal to noise ratio (SNR) and simple fabrication process. We will describe the principle and the performance of the sensor. In addition, we will present some X-ray images of a human subject obtained with this sensor. Dynamic range of the sensor covers most of the exposure range for radiography. SNR is limited almost only by the X-ray photon noise. MTF is sufficient for digital chest radiography. X-ray images have good contrast. The experimental results and obtained images show that the brand new sensor has great advantages for replacing X-ray film. The simple fabrication process of the sensor promises high productivity and low cost of DRS.

Efforts to integrate projection radiography into the digital environment have, to date, required signal degrading steps. The purpose of this study was to compare new directly acquired digital projection radiographic images to conventional film screen images. Fifty paired images (25 chest and 25 abdomen) were obtained under identical conditions and at comparable exposures using a new digital system and a conventional 200 speed film-screen system. This new direct x-ray converting full field 14 X 17 inch detector (Sterling Imaging) uses selenium coupled to a 2560 X 3072 thin film transistor array with a pixel pitch of 139 microns. The detector was easily retrofitted to existing radiographic equipment. After applying appropriate algorithms to obtain images that were comparable in gray scale appearance to conventional film, the 14 bit digital images were printed at full resolution (8 bit) on laser film. Detail evaluation of these paired images under identical viewing conditions, using standardized protocols that were formulated prior to imaging, was performed by three experienced radiologists for each body area. The hard copy clinical digital images were judged by all of the expert panel of radiologists to be superior or equivalent to their paired conventional film screen study (t-value confidence level of 10^-6 for chest and .03 for abdomen).

A new high speed/high resolution X-ray detector called the XEBIT (X-ray sensitive Electron Beam Image Tube) is currently under development at the University of Connecticut Health Center. This large area (9' diameter) direct conversion detector is based on an X-ray photoconductor called thallium bromide. The device utilizes cathode ray tube technology to provide a 30 frame per second raster scanned electron beam to both charge and readout the photoconductor. Thallium bromide is a high Z material with a linear attenuation coefficient of 28.11 cm^-1 at 60 kev. This high stopping power results in a quantum efficiency of 57% at 60 kev for 300 micron thick layers. Thallium bromide is a very good X-ray photoconductor that requires 6.5 ev to create an electron-hole pair. For 60 kev photons, this results in a gain 9230 per absorbed photon. With a hole-mobility lifetime product of 1.5 X 10^-6 cm²/volt, good charge collection can be achieved at reasonable field strengths. Thallium bromide has a very high band gap of 2.7 ev and a dielectric constant of 33. Its resistivity, which is 5 X 10⁹ ohm-cm at room temperature, is dominated by ionic conductivity. Fortunately, ionic conductivity has a strong temperature dependence that can be significantly reduced with moderate cooling to -25 degrees centigrade. The XEBIT uses thallium bromide as an X- ray photoconductor in a vidicon type image tube. Its principals of operation are very similar to the standard light sensitive vidicon that were utilized extensively in the commercial television industry. A scanning electron beam charges the TlBr surface, with respect to the positively biased front surface, providing the necessary electric field across the photoconductor for charge transport. X-rays then penetrate the window and are absorbed by the thallium bromide. The absorbed photons generate large numbers of electron-hole pairs due to the high conversion gain. Electrons drift under the electric field to the positive bias electrode and the holes drift to the vacuum surface and annihilate stored charge. This results in an image dependent charge pattern on the vacuum surface of the photoconductor. A subsequent scan of the photoconductor generates the capacitively coupled signal by replacing the annihilated electrons. The XEBIT utilizes well-developed display tube technology to provide a very cost effective alternative to image intensifier and screen/film based systems. The XEBIT is currently under development as a replacement for X-ray Image Intensifiers in medical imaging applications. The first devices are 9 inch prototypes designed to be no larger than standard intensifiers. It will replace the image intensifier/optics/video camera with one direct conversion device. The XEBIT suffers from no veiling glare and has far superior contrast resolution with over 50 percent modulation at 5 line pairs per millimeter. The XEBIT is capable of full field imaging as well as under scanning to view smaller regions with higher detail.

The image quality of Gd₂O₂S:Tb phosphor screens used in a flat-panel photodiode array system is examined. The presampled Modulation Transfer Function (MTF), the Normalized Noise Power Spectrum surface (NNPS) and the resulting Detective Quantum Efficiency (DQE) are discussed for a variety of screens used in such a system. A technique for extracting the limiting DQE for a system with significant electronic noise is described. This allows for the examination of the imaging performance of the X-ray converter (or phosphor screen) and removes issues of photodiode array and readout electronics performance. It is shown that depending on the metric being used to judge the imaging performance (i.e., MTF or DQE), it is possible to design a more optimal screen than those currently available for use in screen-film imaging. Evidence is also presented for a significant degradation of the DQE at higher spatial frequencies due to the variation in the light spread MTF through the depth of the screen.

New radiological imaging techniques have to be capable of high image quality, high acquisition speed, compactness, and versatility in operation. We have been investigating a flat panel imager based on the direct conversion of x-ray energy to an electric charge signal in a layer of amorphous selenium (a- Se) and an active matrix real-time self scanning read out, that is expected to achieve these goals. Although prototype imagers have demonstrated both real-time acquisition and the quality necessary for radiography, systematic investigations of the photoconductive properties of a-Se were undertaken to understand and extend the photoconductor range of operation and facilitate the designs for specialized radiological applications, such as fluoroscopy. The method of pulse height spectroscopy was adapted to accurately measure the reciprocal photoconductive gain W_+/- (eV/e^-h⁺) as a function of the incident photon energy (epsilon) and the applied field E. Improvements were taken until spectral peaks could be well resolved. The accuracy of the measurements was increased by calibration with a Si-PIN diode with a known W_+/- . The measured values of W_+/- showed that gain increases of a factor of 2.6 could be realized at the highest fields. The spectral widths were corrected to give the gain fluctuations and then used to calculate the contribution to imaging noise, i.e. the Swank factor. The imaging noise contribution was shown to be negligible, in agreement with previous calculations. The data suggests a new model of charge recombination in low mobility semiconductors which can be used to calculate W_+/- at different fields and x-ray spectra for the full range of radiological applications.

Digital x-ray imaging techniques of today require electronic detectors that can be applied to all modalities of medical imaging. This paper presents work showing that selenium, when used as a direct converter, can be competitive with present day scintillator technologies targeting mammographic, radiographic and fluoroscopic applications. In this work, supporting results are presented on the dark currents, lag and ghosting effects and also on the theoretical and experimental x-ray absorption and sensitivity of selenium layers. Measurements were carried out on suitably alloyed selenium layers where the electronic transport properties have been optimized. Measured values for dark currents were below 100 pA/cm² at operating fields up to 20 V/micrometer. Experimental measurement of the intrinsic lag in selenium has shown it to be less than 0.5% after 30 milliseconds under a dose of 50 mR at 55 keV mean beam energy, which is very low compared with present day image intensifiers. Similar measurements on ghosting, using multiple radiographic pulses, indicate that the magnitude of the ghost image after a few seconds is around 2000 electrons, which is comparable to the electronic noise of most read out systems. Measured sensitivity of a 200 micrometer selenium layer under a mammographic spectrum was around 230 pC/mR/cm² at an operating field of 20 V/micrometer, which is significantly higher than that reported for competing technologies. Sensitivity for 1000 micrometer selenium was also measured with an 80 kVp spectrum and 20 mm Al filtration and was found to be around 3400 pC/mR/cm² which is in close agreement with theoretically calculated values. Theoretical estimations for MTF and DQE are also given to assess the potential imaging performance of a selenium-based detector for various applications.

An amorphous silicon medical imaging system designed to operate in both radiographic and fluoroscopic modes is described. Images of medical phantoms are presented for both modes of operation. MTF and DQE measurements are also presented. The effect of recursive filtering on the DQE performance of the system operating in fluoroscopic mode is discussed.

A VLSI readout chip for a solid-state hybrid pixel detector for two-dimensional real-time imaging is part of a project with the goal to construct a large-area digital detector with good efficiency for X-rays in the energy range 20 - 70 keV. A new circuit concept with integrated analog counters in each cell makes the detector suitable for high-intensity dynamic imaging. The aim is to acquire 12-bit images with a speed of up to 100 frames per second. The detector concept also includes single-photon counting and an adjustable threshold. The size of the pixel is a trade-off between spatial resolution and amount of implemented logic, and is chosen to 270 micrometer X 270 micrometer. A 8 X 8-channel prototype readout chip, (ANGIE), with a pixel size of 250 micrometer X 250 micrometer, has been fabricated and tested. The results from functional tests are presented.

Thin Film Transistor (TFT) array technology is presented for Digital X-ray Sensors in Direct Radiography applications. Circuit simulations were performed to optimize the design of the TFT array. The sensor array uses a combination of a mushroom electrode with a high fill factor of 86% and a polymer passivation dielectric to minimize column capacitance and improve signal-to-noise ratio. A 14 in. X 8.5 in. sensor array with 1536 X 2560 pixels was developed using this technology. The TFT arrays are processed entirely in Class 1 clean room environments to eliminate line defects and minimize pixel defects. The best 14 in. X 8.5 in. panels have exhibited fewer than 0.001% pixel defects, as detected during in process testing prior to Se coating. In typical image quality comparisons with conventional X-ray film/screen combinations, the digital X-ray sensor exhibited equal or better performance than film-screens. Clinical studies were also conducted. Radiologists concluded that diagnostically significant projection radiographic images can be produced with the new digital X-ray sensor that are equivalent or superior to conventional film/screen images at the same X-ray exposures. The detector recently received FDA approval.

We present an update on a novel direct digital X-ray imaging device and system. The system comprises a mosaic of hybrid solid state semiconductor devices removably mount onto a master plane covering an imaging surface of any desirable shape and size. Each imaging device comprises a pixel semiconductor detector flip-chip joined to a CMOS ASIC. Monolithic CdZnTe and Si pixel detectors with dimensions 12.2 X 4.2 mm² and 18.9 X 9.6 mm² have been implemented with a pixel pitch of 35 micrometer. Each circuit on the ASIC, corresponding to a detector pixel, is capable of accumulating thousands of X-rays in the diagnostic energy spectrum with high efficiency (CdZnTe) and user accumulation times ranging from just a few ms to a few s. Individual, removable tiles are combined in a mosaic providing continuous large area imaging with no inactive regions. This tiling approach allows for cost efficient replacement of defective tiles. The packaging delivers a compact, lightweight, portable cassette whose thickness is around 2.0 cm. The basic hybrid detector design and tiling scheme are generic and may be used in mammography, conventional radiography and fluoroscopy. A special tiling scheme has been designed for use in intraoral imaging. We present our measured Modulation Transfer Function (MTF) and Detective Quantum Efficiency (DQE). Images taken with hard objects, phantoms and soft tissue further demonstrate system functionality and provide a comparison with radiographic film and CR plates. The first application of the new technology is intended for the field of dental imaging, mammographic biopsy and other small area medical applications (approximately 10 - 30 cm² imaging area) as well as Small Area Non Destructive Testing.

In this paper, the detected signal-to-noise was measured and related to the tube current (mA) setting. The line spread function amplitude (LSF) dependence on drift distance of a 3 mm thick detector, for 100 kVp, 100 mA, with an applied electric field of 50 V/mm, and 100 V/mm, were measured. In addition, the dependence of the modulation transfer function [MTF(f)] of the x-ray detector system on the applied bias voltage has been experimentally determined. The experimental setup, although is not offered for large field-of-view imaging applications, offers capabilities for feasibility studies, research and evaluation of the temporal response and noise characteristics of a Cd_1-xZn_xTe detector, for fast digital radiographic and CT applications. The experimental results indicate that Cd_1-xZn_xTe detectors exhibit both a high signal-to-noise ratio and linear response, as well as a good spatial resolution within the diagnostic energy range. Tor system improves both with increasing applied bias voltage and decreasing detector thickness. A study is in process aimed at improving the spatial resolution of the x-ray system by suitable optimization of the system geometry as well as the system temporal response.

The performance of indirect-detection flat-panel imagers incorporating CsI:Tl x-ray converters is examined through calculation of the detective quantum efficiency (DQE) under conditions of chest radiography, fluoroscopy, and mammography. Calculations are based upon a cascaded systems model which has demonstrated excellent agreement with empirical signal, noise- power spectra, and DQE results. For each application, the DQE is calculated as a function of spatial-frequency and CsI:Tl thickness. A preliminary investigation into the optimization of flat-panel imaging systems is described, wherein the x-ray converter thickness which provides optimal DQE for a given imaging task is estimated. For each application, a number of example tasks involving detection of an object of variable size and contrast against a noisy background are considered. The method described is fairly general and can be extended to account for a variety of imaging tasks. For the specific examples considered, the preliminary results estimate optimal CsI:Tl thicknesses of approximately 450 micrometer (approximately 200 mg/cm²), approximately 320 micrometer (approximately 140 mg/cm²), and approximately 200 micrometer (approximately 90 mg/cm²) for chest radiography, fluoroscopy, and mammography, respectively. These results are expected to depend upon the imaging task as well as upon the quality of available CsI:Tl, and future improvements in scintillator fabrication could result in increased optimal thickness and DQE.

Results of an investigation into the limiting spatial resolution of a flat-panel amorphous silicon (a-Si:H) X-ray imaging system are reported. The system was comprised of a 127 micrometer pixel pitch a-Si:H array used in conjunction with an overlying Gd₂O₂S:Tb (GOS) phosphor screen. The pre- sampled modulation transfer function (psMTF) of the system was measured at diagnostic X-ray energies and compared to the value predicted from a knowledge of the spatial resolution of the individual system components. A reproducible drop in the measured psMTF is seen at low spatial frequencies. Measurements of the magnitude of X-ray backscatter from the array substrate, along with the results of a theoretical model for K-fluorescence X-ray scatter, indicate that a significant fraction of this low-frequency drop is due to K-fluorescence from heavy elements in the glass substrate of the array. This K-fluorescence may be excited directly by primary X-rays that penetrate the overlying phosphor and interact in the glass, or by gadolinium K-fluorescence X-rays that escape from the phosphor into the glass. The measurements indicate that the spatial resolution of such an X-ray imaging system may be improved by the use of a substrate containing as low a concentration of heavy elements as possible.

The photostimulated luminescence (PSL) characteristics of alkali halide phosphors, such as KCl:Eu, KBr:Eu and KCl_xBr_1-x:Eu phosphors, as new storage photostimulable phosphor materials for the imaging plate in computer radiography are investigated. Intense 420 nm PSL is observed in X-ray-irradiated KCl_xBr_1-x:Eu phosphor. The PSL intensity increases linearly with increasing X-ray irradiation dose over the wide range, showing a good storage material for the imaging plate in computed radiography. The KBr:Eu phosphor exhibits excellent fading characteristics, which showing that the KBr:Eu phosphor is useful as a material for medical imaging utilizing PSL phenomenon. The mechanism of the fading is also discussed.

A new dual screen-dual film mammography combination was constructed which made of two phosphor screens and two films loaded into a single x-ray cassette. The screens and films were combined so that a single emulsion film (Film #1, Kodak Min-R E film) was placed in direct contact with the phosphor side of the first screen (Screen #1). Screen #1 was made of the Kodak Min-R phosphor (34.0 mg/cm² Gd₂O₂S:Tb) coated on a 4 mil transparent Mylar backing. A double emulsion film (Film #2, Kodak T-Mat G film) was sandwiched between Screen #1 backing and the phosphor side of the second screen (Screen #2). Screen #2 was a Kodak Insight ME screen that has a Gd₂O₂S:Tb coating weight of 43.1 mg/cm² and a reflective coating between its phosphor layer and support layer. The relative sensitometric responses and spatial resolution properties of the two films were measured as a function of x-ray tube potential (kVp). A series of a contrast-detail phantom images was obtained by varying the x- ray exposure level at 28 kVp. An observer performance study was subsequently carried out to investigate the low contrast performance using the dual screen-dual film combination. The effective speeds of the two films differed by a factor of 2.12 to 2.67 between 25 to 30 kVp. Film #2 contrast was a factor of two or greater than Film #1 in the range where Film #1 optical density values were under 0.7. Measured MTF curves from digitized edge images showed that Film #1 MTF performance was comparable to a standard Kodak Min-R screen-Min-R E film combination. The limiting spatial resolution was found to be 19.5 lp/mm for Film #1 and approximately 11 lp/mm for Film #2. Observer performance studies showed that the threshold contrast in detecting small (less than 10 mm) breast lesions could be up to a factor of two lower on Film #2 images when regions of interest are underexposed on Film #1 images.

An investigation was made of the key determinants of the detective quantum efficiency at zero spatial frequency [DQE(0)] of a CsI:Tl scintillator based scanning slot x- ray detector for digital mammography. The slot x-ray detector was made of a prismatic type thallium activated CsI scintillator (150 micrometer thick) optically coupled to CCDs by fiber optical image guides. Monte Carlo calculations were performed to generate the CsI:Tl scintillator Swank factor on the basis of the energy deposition from pencil beam x-ray sources and light transmission within the CsI:Tl scintillator. A theoretical expression for the detector DQE(0) was obtained which was used to investigate the detector imaging performance as a function of x-ray energy, x-ray exposure, CCD electronic noise level, and optical coupling efficiency of the fiber optic image guide. The Swank factor of the CsI:Tl scintillator was close to unity at x-ray energies below Iodine K-edge (33.2 keV), but decreased to approximately 0.8 at higher x-ray energies up to 40 keV. DQE(0) of the slot x-ray detector was approximately 75% at 15 keV but decreased to approximately 40% at 30 keV. Optimum DQE(0) performance of the slot x-ray detector was generally obtained at a detector x-ray exposure level above approximately 5 to 10 mR and an electronic noise level below approximately 50 electrons rms. A drop in the optical coupling efficiency of the image guide from 1.0 to 0.3 reduced the detector DQE(0) from approximately 75% to approximately 55% in the mammography x-ray energy range. The key finding in this study is that the choice of the x-ray energy has a major impact on the DQE(0) of a CsI:Tl scintillator based slot x-ray detector. Since the x-ray photon energy also affects x-ray tube loading, mean glandular dose and subject contrast, the choices of optimal x-ray spectra from current mammography x-ray tubes require further investigation.

A thorough assessment of the performance of an imaging system includes a measurement of the detective quantum efficiency, DQE(f). One of the terms which is required to calculate DQE(f) is the x-ray fluence (photons/mm²). Diagnostic x-ray systems make use of a wide variety of x-ray spectra, with different kV's, waveforms, and filtration. However, most investigators do not have the equipment to measure the x-ray spectrum directly. The determination of x-ray quantum fluence, which is strongly dependent upon the spectra, is therefore left to approximation. In this paper, a wide variety of x-ray spectra for both mammography (with Mo, Rh and W anodes and 18 - 40 kV) and general diagnostic radiography (W anode, 40 - 140 kV) with different patient (Plexiglas) thicknesses were modeled (using a technique which interpolates measured x-ray spectra), and the quantum fluence was tabulated. Five extensive tables (including the calculated HVL's) are provided to allow investigators to interpolate fluence values appropriate to the x-ray system under study. It is hoped that usage of these tables will prove useful to investigators in their assessment of system performance, and perhaps better consistency between measurements can be achieved.

Following procedures used to simulate the image sharpness along the radiation field based on X-rays geometric exposure developed in previous work, here we describe another computer simulation procedure intended to evaluate the influence of any recording system as radiographic film or screen-film combinations, on the image sharpness in mammography. In this current work we take into account the parameters from the recording system besides the radiation projection from the focal spot in order to yield a simulated image on the computer screen relative to the expected image to be obtained in actual conditions with a singular recording system for a singular mammography equipment. The focal spot sizes in all field locations, as well as the respectives intensity distributions, the sensitometric curve for a radiographic film or for a screen-film combination, and also the screen intensifying factor, conversion efficiency, absorption factor and emission spectrum were used as input parameters for the simulation. Simulated images were compared to those obtained with actual mammographic equipment, by using a resolution phantom, and both types of images were in good agreement. The main advantage of this procedure will be the possibility of predicting the image sharpness characteristics for any mammography equipment with any type of recording system without exposure tests.

Quality control practices in screen/film mammography are inadequate and not necessarily suitable for digital mammography due to differences in spatial resolution, contrast, and artifacts. And screen/film exposure techniques are not useful in determining the correct digital techniques. This study has produced changes and additions to mammography quality control appropriate for digital systems, and has revealed necessary changes in exposure to optimize digital image quality. Quality control has been studied for the TREX whole breast digital system with a phantom designed to test each CCD individually for SNR and calcification conspicuity. In addition, white fields were compared at different time intervals to determine the necessary frequency of recalibration. Optimal exposure techniques were determined by varying kVp, mAs, and filter in order to maximize SNR and calcification conspicuity while minimizing mean glandular dose. Variations in the white field were found to necessitate weekly recalibrations. Increasing the kVp or mAs or both was found to improve SNR and calcification conspicuity, and if rhodium filtration instead of molybdenum was used, mean glandular dose was lowered to mandated levels with no loss in SNR.

The objective of this research was to compare a Fischer MammoVision/MammoTest and a LoRad DSM digital biopsy machine using the Computer Analysis of Mammography Phantom Images (CAMPI) methodology. This study reports on analysis of the 4 largest microcalcification groups (M1, M2, M3 and M4) and the largest nodule (N1) in a mammography accreditation phantom on images acquired at 26 kVp and different mAs values on the two machines. Both machines were linear in response but the MammoTest was more sensitive (i.e., it yielded a larger gray- scale value for a given x-ray technique). However, even after correcting for this difference, the CAMPI noise measure was substantially smaller for the LoRad than the MammoTest over the range of mAS values studied. Similarly, the CAMPI signal- to-noise-ratio and correlation measures were higher for the LoRad than the MammoTest over the same range of mAs, especially for the larger objects (M1/M2 and N1). For the smaller specks in M3/M4 somewhat closer performance was observed. The overall differences are attributed to better contrast/noise performance of the LoRad which appear to outweigh its lesser resolution capability. Our results are in agreement with earlier physical and psychophysical measurements using different methodologies. This work also describes better predictive models (i.e., fits) to describe the variation of all CAMPI measures with mAs at constant kVp. For example, the noise measure was fitted to a function that included physically reasonable sources of noise e.g., dark noise and detector gain fluctuations, in addition to the usual quantum noise. These fits can be used to summarize machine performance and to predict dependencies on other variables (e.g., exposure or dose) that are related to the mAs.

This work is about the development of a computer-based instrument intended to be applied to quality assurance of mammography systems. It was designed in order to obtain information about kVp, mA, exposure time, dosage, waveform and other important operational characteristics of the radiographic equipment by means of few singular exposures. In addition, the measurement of the focal spot sizes is provided, as in the field center, as in any other location. Indeed the system is designed to determine automatically the position of the sensor relative to the field center, independent on the place where it is positioned on the equipment table. Thus, the instrument is composed by a system which determines automatically the position of the test object relative to the field center, and it is based on silicon sensors coupled to an electronic amplifier system, and to a notebook computer. Calculations made by the software, developed specifically for this application, can correct automatically the shifts of positioning of the test objects relative to the field center. By using the field characteristics equations, it is able to determine also the target angulation on the anode, and intensity distributions characteristics due to the Heel effect. The notebook screen can display, thus, information about the operational characteristics of the radiological system, as well as a diagram of the effective focal spot characteristics along the radiation field. In tests performed with a mammography equipment, the instrument was able to determine the operational conditions with an average accuracy of +/- 1% and the average error of the focal spot sizes measurements was as low as 10 micrometer.

Two main phenomena degrade the quality of a mammogram: the blur induced by X-rays scattering and a loss of contrast due to beam hardening. In this paper, we propose an original approach to restore the mammographies from these degradations. They are due to the physical phenomena occurring in the radiographic image formation process. Our objective has been to construct a physical model describing accurately the phenomena while being tractable to enable an inversion and therefore to construct a corrected mammography. In fact, we show that, in reason of the specific protocol for realizing a mammographic exam, and the composition of the breast in mainly two components (fat and glandular tissue), the observed values can be related to the thickness of glandular tissue crossed by the X-rays. This relation is nonlinear, and expresses the different phenomena taking place in the acquisition process. An adapted inversion scheme enables to build up a map of the thicknesses of glandular tissue. This representation enhances significantly the mammograms. As the approach only relies, on one hand on priors deduced from the acquisition geometry and from the breast composition, and on the other hand on a physical description of the acquisition process, it does not create artifacts that may alter the physician's diagnosis.

We have studied the influence of incorrect placement, such as the lateral decentering, the inclination and the x-ray source out of the focused distance, of the focused grid on the image formation in a computer simulation. We have applied Monte Carlo methods to determine the absorption and scattering of photons inside water phantoms and traced the radiation transmitted from the phantom through the focused grid. We have examined the dependence of the transmittance of primary radiation and that of the transmittance of scattered radiation on various grid parameters and radiographic conditions. The grid exposure factor and the contrast improvement factor as well as the primary radiation loss were used for comparison of the relative performance of the grid, and relationship among these factors was studied. We found that the lateral decentering and the inclined grid degraded the image remarkably. The higher the grid ratio is, the more sensitively the misplacement of the grid degrades to the image.

An automatic ROI tracking system for the application of ROI fluoroscopy is designed. The methods for processing in real- time, generating binary masks, equalization of the display and the scheme for the image analysis to track ROI size are described. The initial results of the system tested on a torso phantom are reported and discussed. The system results in the appearance of an almost completely equalized real-time image and the method can be incorporated into x-ray fluoroscopy units with minimum modification. With the technology, ROI fluoroscopy can be applied to interventional or diagnostic procedures for practical patient dose reduction.

Fluoroscopic images are degraded by scattering of x-rays from within the patient and by veiling glare in the image intensifier. Both of these degradations are well described by a response function applied to the primary intensity. If the response function is known, than an estimate of the primary component of the image can be computed by applying the inverse operation. However, the response function is actually variable, with dependence on such factors as patient thickness and imaging geometry. We describe a technique for estimating a parameterized response function so that a good estimate of the subject density profile can be recovered even if the response function parameters are not known in advance. Our method uses a partially absorbing filter with spatially varying density as a reference object which enables us to compute good estimates of the parameterized response function. We use simulated images to evaluate our method for a wide range of conditions. Our simulation results show that this technique can greatly reduce densitometric errors in fluoroscopic images.

Rossmann proposed that the Wiener spectrum of the quantum mottle was proportional to the square of the modulation transfer function (MTF) of the screen-film system. On the other hand, Lubberts pointed out that the shape of the Wiener spectrum of the quantum mottle depended on the sum of the squares of the MTFs for different depths in the screen phosphor layer, rather than the square of the sum of the MTFs for the different depths, i.e., the square of the MTF of the screen-film system. The purpose of this study is to experimentally investigate the proportionality between the Wiener spectrum of the quantum mottle and the square of the MTF of the screen-film system using two screen-film systems having different screen thicknesses. For the purpose, we determined correction factors for the square of the MTF of the screen-film system in the Wiener spectrum of the quantum mottle, that is, the ratios of the sums of the squares of the MTFs for different depths to the squares of the MTFs of the screen-film systems so that the theoretical Wiener spectral values of the screen mottle fitted the experimental values. For the thin screen, the correction factors were unity for all spatial frequencies, that is, the Wiener spectra of the quantum mottle were proportional to the square of the MTF of the screen-film system. On the contrary, for the thick screen, the factor increased with the spatial frequency, that is, the Wiener spectra were proportional to the sum of the squares of the MTFs. Therefore, we can conclude that the relation between the Wiener spectrum of the quantum mottle and the MTF of the screen-film system, for thin screen, agrees with Rossmann's theory, whereas, for thick screen, agrees with Lubberts' theory.

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.

The Kinestatic Charge Detector (KCD) is an electronic digital strip beam x-ray detector which has been shown to possess a high detective quantum efficiency, good spatial resolution, and good scatter rejection. We have investigated its use as a dual-energy x-ray detector detector, which involves the acquisition of two images with different mean x-ray energies that can be reconstructed using a suitable algorithm to form images of two basis materials such as bone and soft tissue. Dual-energy imaging with a single exposure may be performed with a KCD by segmenting its x-ray collection region into front and back regions. The lower x-ray photons will then be preferentially absorbed in the front region. Computer simulations were performed to evaluate a segmented KCD's ability to reconstruct various combinations of Plexiglas and aluminum. Actual experimental data were also taken for various Plexiglas and aluminum combinations with a non-imaging research KCD. The suitability of using analytic calibration functions as decomposition algorithms for aluminum and Plexiglas basis material images was investigated. Fits were performed for the computer simulations using the high-energy and low-energy data, with and without the addition of noise. Similar fitting techniques were used with the experimental KCD data. A true rms accuracy of 150 micrometer for aluminum and 500 micrometer for Plexiglas was obtainable from fits for the computer simulated data, even with the addition of noise. The experimental data taken with the non-imaging KCD yielded rms errors of approximately 250 micrometer for aluminum and 1000 micrometer for Plexiglas, comparable to simulated noisy data. We conclude that suitable decomposition algorithms exist for a segmented dual-energy KCD to be able to reconstruct aluminum and Plexiglas material thicknesses to an accuracy sufficient for clinical diagnosis in chest radiography.

Conventional means of diagnosiing and assessing the progression of osteoporosis, including radiographic absorptiometry and quantitative CT, are directly or indirectly dependent upon bone density. This is, how ever, not always a reliable indicator of fracture risk. Changes in the trabecular structure and bone mineral content (BMC) are thought to provide a better indication of the change of spontaneous fractures occurring. Coherent-scatter CT (CSCT) is a technique which produces images based on the low angle (0 - 10 degrees) x-ray diffraction properties of tissue. Diffraction patterns from an object are acquired using first-generation CT geometry with a diagnostic x-ray image intensifier based system. These patterns are used to reconstruct a series of maps of the angle dependent coherent scatter cross section in a tomographic slice which are dependent upon the molecular structure of the scatterer. Hydroxyapatite has a very different cross section to that of soft tissue, and the CSCT method may, therefore, form the basis for a more direct measure of BMC. Our original CSCT images suffered from a 'cupping' artifact, resulting in increased intensities for pixels at the periphery of the object. This artifact, which is due to self-attenuation of scattered x rays, caused a systematic error of up to 20% in cross-sections measured from a CT image. This effect has been removed by monitoring the transmitted intensity using a photodiode mounted on the primary beam stop, and normalizing the scatter intensity to that of the transmitted beam for each projection. Images reconstructed from data normalized in this way do not exhibit observable attenuation artifacts. Elimination of this artifact enables the determination of accurate quantitative measures of BMC at each pixel in a tomograph.

We investigate the use of the kinestatic charge detector (KCD) together with the multi-level scheme algebraic reconstruction technique (MLS-ART) for computer tomography (CT) reconstruction, to be used in position verification in radiotherapy. The KCD offers very good contrast resolution, which is especially useful given the low number of projections we are aiming at. We present the images reconstructed using a head phantom (Rando-phantom) using a total of 95 projections, and a standard low contrast CT phantom using 63 projections. The reconstruction was carried out using MLS-ART technique, in this technique satisfactory images are generally obtained after one or two iteration, which in effect makes ART a noniterative algorithm. We also present the CT images obtained using the back projection technique for comparison purposes.

The prime motivation of this work is to devise a projection algorithm that makes the Algebraic Reconstruction Technique (ART) and related methods more efficient for routine clinical use without compromising their accuracy. While we focus mostly on a fast implementation of ART-type methods in the context of 3D cone-beam reconstruction, most of the material presented here is also applicable to speed up 2D slice reconstruction from fan-beam data. In this paper, we utilize the concepts of the splatting algorithm, which is a well known and very efficient voxel-driven projection technique for parallel projection, and devise an extension for perspective cone-beam projection that is considerably more accurate than previously outlined extensions. Since this new voxel-driven splatting algorithm must make great sacrifices with regards to computational speed, we describe a new 3D ray-driven projector that uses similar concepts than the voxel-driven projector but is considerably faster, and, at the same time, also more accurate. We conclude that with the proposed fast projection algorithm the computational cost of cone-beam ART can be reduced significantly with the added benefit of slight gains in accuracy. A further conclusion of our studies is that for parallel-beam reconstruction, on the other hand, a simple voxel-driven splatting algorithm provides for more efficient projection.

An image intensifier-based rotational volume tomographic digital angiography (VTDA) imaging system has been constructed. The system consists of an x-ray tube and an image intensifier that are separately mounted on a gantry. This system uses an image intensifier coupled to a charge coupled device (CCD) camera as a two-dimensional (2D) detector so that a set of 2D projections can be acquired for a direct three- dimensional (3D) reconstruction. This paper presents the results of the preliminary animal studies using the volume tomographic angiography system with intravenous (IV) injections. The results from the animal studies demonstrate that VTDA only needs a single IV injection of contrast and volume scanning to provide adequate image contrast and resolution for quantitative assessment of vascular anatomy.

With the rising prominence of filmless and quantitative imaging technologies in modern cardiac catheterization laboratories, new approaches to assessing angiographic image quality (AIQ) are required. This paper describes a system of radiographic phantom modules and associated digital image analysis software for assessing and tracking the inherent low contrast AIQ of a particular image chain as well as its suitability for quantitative coronary arteriography (QCA) analysis. The system uses two phantom modules in conjunction with a patient simulation apparatus. Computer software operates on digital radiographic images of the phantoms and extracts a number of parameters characterizing the imaging performance of the particular image chain. An arterial phantom module is used to evaluate angiographic imaging performance relating to QCA and an orthogonal array of small bronze ball bearings is used to derive AIQ parameters that complement standard manual measurements. In addition to passively monitoring AIQ, this new approach has the benefit of enabling active compensation for variability in AIQ and its effect on QCA results from one image chain to the next.

With the advent of high heat loading capacity x-ray tubes, high frequency inverter type generators, and the use of spectral shaping filters, the automatic brightness/exposure control (ABC) circuit logic employed in the new generation of angiographic imaging equipment has been significantly reprogrammed. These new angiographic imaging systems are designed to take advantage of the power train capabilities to yield higher contrast images while maintaining, or lower, the patient exposure. Since the emphasis of the imaging system design has been significantly altered, the system performance parameters one is interested and the phantoms employed for the quality assurance must also change in order to properly evaluate the imaging capability of the cardiovascular imaging systems. A quality assurance (QA) phantom has been under development in this institution and was submitted to various interested organizations such as American Association of Physicists in Medicine (AAPM), Society for Cardiac Angiography & Interventions (SCA&I), and National Electrical Manufacturers Association (NEMA) for their review and input. At the same time, in an effort to establish a unified standard phantom design for the cardiac catheterization laboratories (CCL), SCA&I and NEMA have formed a joint work group in early 1997 to develop a suitable phantom. The initial QA phantom design has since been accepted to serve as the base phantom by the SCA&I- NEMA Joint Work Group (JWG) from which a comprehensive QA Phantom is being developed.

An enhancement technique for phase dependent image contrast is proposed. Because the method can enhance inherent phase contrast it is suited for susceptibility imaging and flow imaging where intravoxel phase is a source of image contrast. In this paper, applying external phase in the voxel enhances phase contrast. The external phase is generated by a tailored RF pulse so that one can control the phase contrast and even produces phase only contrast. Signal intensity due to both inherent phase and external phase is analyzed and the proposed technique is applied to a susceptibility effect only imaging and a flow effect only imaging. To verify the proposed technique, computer simulations are performed and their results are given.

This paper presents a simulation methodology which includes all steps of Magnetic Resonance Imaging (MRI), from precession of magnetic moments to reconstruction of the image. For modeling the operation of RF pulses, well known solutions to the Bloch equations are used. To simulate the effects of arbitrary shape RF pulses, a new physical based method is presented. Intravoxel spin dephasing is simulated and partial volume averaging and noise are included in the simulation procedure. To test and validate the algorithm, MR images of a Quality Control (QC) phantom and a human brain are simulated by the proposed approach. The resulting images are compared to the corresponding actual MR images. Impulse response of a spin echo sequence for different conditions such as different locations for the impulse, different amounts of static magnetic field nonuniformity, and different number of spins in the voxel are generated by the simulation algorithm.

High demands are made of the gradient systems of MRI scanners during acquisition of echo-planar images. These may occasionally lead to spurious noise added to the received data in the form of short duration, high intensity 'spikes.' If these spikes are not caught prior to image formation, a severely degraded image may be formed. In the acquisition of functional MRI data this degradation may prohibit subsequent analysis of the data for identification of activated brain regions. We have devised several algorithms to isolate and remove these spurious spikes prior to the Fourier transform, and in one example show that subsequent fMRI analyses are possible. By contrast, the analysis is either impossible or produces differing results without this processing step.

Due to the dynamic nature of brain studies in functional magnetic resonance imaging (fMRI), fast pulse sequences such as echo planar imaging (EPI) and spiral are often used for higher temporal resolution. Hundreds of frames of two- dimensional (2-D) images or multiple three-dimensional (3-D) images are often acquired to cover a larger space and time range. Therefore, fMRI often requires a much larger data storage, faster data transfer rate and higher processing power than conventional MRI. In Mercury Computer Systems' PCI-based embedded computer system, the computer architecture allows the concurrent use of a DMA engine for data transfer and CPU for data processing. This architecture allows a multicomputer to distribute processing and data with minimal time spent transferring data. Different types and numbers of processors are available to optimize system performance for the application. The fMRI reconstruction was first implemented in Mercury's PCI-based embedded computer system by using one digital signal processing (DSP) chip, with the host computer running under the Windows NT^R platform. Double buffers in SRAM or cache were created for concurrent I/O and processing. The fMRI reconstruction was then implemented in parallel using multiple DSP chips. Data transfer and interprocessor synchronization were carefully managed to optimize algorithm efficiency. The image reconstruction times were measured with different numbers of processors ranging from one to 10. With one DSP chip, the timing for reconstructing 100 fMRI images measuring 128 X 64 pixels was 1.24 seconds, which is already faster than most existing commercial MRI systems. This PCI-based embedded multicomputer architecture, which has a nearly linear improvement in performance, provides high performance for fMRI processing. In summary, this embedded multicomputer system allows the choice of computer topologies to fit the specific application to achieve maximum system performance.

We have made a detailed study of the response of LSO detectors to 511 keV (gamma) -rays. The LSO, discovered recently, has a density greater than BGO, small decay time, and high light output. As such, it should have an overall behavior better than that of BGO. We have modeled a (gamma) -ray detector using an LSO crystal of rectangular cross-section attached to a photomultiplier tube (PMT). We used our PET simulation package to study the energy resolution, efficiency, and timing resolution for various crystal sizes and various energy thresholds. The simulation takes into account the interactions of (gamma) -rays in the crystal via Compton and photoelectric effects, the production and transport of scintillation photons, the productions of photoelectrons in the PMT and the anode signal formation. We have estimated the efficiency versus energy threshold for various lengths of the LSO crystal and we find the for 400 keV threshold this efficiency is large even for 2 cm crystals and comparable to that of BGO. We also estimated the timing resolution (FWHM) versus crystal length for various energy thresholds. The timing resolution is comparable to that of CeF₃ detectors. The energy resolution is about 10% (FWHM), which allows one to set the energy threshold fairly high.

Positron Emission Tomography instrumentation has been directed towards improving detector spatial resolution with an emphasis on reduction of detector size. From the early days better resolution has meant ability to easily visualize and delineate regions in brain images in addition to improved accuracy in the measurement of smaller volumes of radionuclide concentration. The major difficulty in achieving both high resolution and high detection efficiency is the requirement for a small detector to fully absorb the 511 keV photon within its volume, thus minimizing Compton scattered photons and also escape into neighboring detector crystals. The latter could be recorded as true events and accepted as part of the data thus leading to mispositioning and consequently loss in spatial resolution. In addition, under certain conditions, identifying the correct pair of detectors in coincidence is ambiguous and an event can be rejected, causing a loss in the efficiency and an increase in the dead time. As individual detector crystals are becoming smaller the contribution of inter-detector scattering needs to be addressed.

We present the evaluation of two Position Sensitive Photomultiplier Tubes (PSPMT), the 1' square Hamamatsu R5900- C8 and the 5' circular Hamamatsu R3292, and discuss their suitability as light sensors in a high resolution small gamma camera for in-vivo radioisotope imaging studies. It has been found that there is a strong dependence between the performance of both tubes and the size of the light distribution at the photocathode. For very narrow distributions (1 mm FWHM) from an optical fiber incident on the R5900-C8, most of the area of the tube is insensitive to position changes. This is related to the size of the anodes and the small spread of electron charge in the dynode multiplication process. For wider distributions (2.2 mm FWHM), we achieved a spatial resolution of 1 mm FWHM. The R5900-C8 has also been coupled to a 5 mm thick CsI(Na) crystal which was scanned across by a 1 mm collimated Tc-99m source (140 keV). The results confirm the importance of light spread at the photocathode. The Hamamatsu R5900-C8 was found to be a promising detector for high resolution radionuclide imaging. The results obtained with the R3292 indicate that a high spatial resolution (less than 2 mm FWHM) is achieved for narrow light distributions (less than or equal to 1 mm FWHM).

A simulation program for x-ray methods is discussed. Computational algorithms for definition of x-ray sources, interactions of x-rays with complex objects and formation of images are developed from first principles. Subject geometries can be accessed from CAD definitions or from CT sets. The principles underlying the image formation process are introduced and images in industrial and medical x-ray applications are displayed.