Although image quality in chest radiography can be improved dramatically with compensating attenuators, it is difficult to match attenuator geometry to that of the patient and to properly position the filter in the x-ray beam. We are developing a system which will fabricate a compensating filter specific for individual patient anatomy and position the filter automatically. A low-dose image is acquired from which the attenuator is designed using a hardwired algorithm. The attenuator then is typed in multiple layers of a cerium oxide material onto a sheet of paper with a dotmatrix printer. Following positioning of the attenuator in the x-ray beam, the final compensated image is acquired with a photographic or electronic detector. Initial studies with our prototype digital beam attenuator have produced images with improved image quality. The improvement occurs because the attenuator allows the entire image to be placed in the linear portion of the film characteristic curve. In addition, decreased scatter fractions and increased signal-to-noise contribute to improved visualization of low-contrast signals behind the most attenuating regions of the patient.

The effect of filter material on the properties of variable-thickness equalization filters for chest radiography was investigated. Filters made of nickel, tantalum, and lead which provided very similar improvements in mediastinal penetration were compared with regard to their scatter fractions, image contrasts and patient exposures. Their performances in these terms were found to be very similar, even though they produce considerably different x-ray spectra. The K edge filters (tantalum and lead) yielded slightly higher image contrast and patient exposure than did the non-K edge filter (nickel). The effect of filter material on contrast and exposure was less than the effect of usual patient-to-patient variation on these quantities, however.

In digital radiography the pixel size used determines the potential spatial resolution of the system. The need for spatial resolution varies depending on the subject matter imaged. In many areas, including the chest, the minimum spatial resolution requirements have not been determined. Sarcoidosis is a disease which frequently causes subtle interstitial infiltrates in the lungs. As the initial step in an investigation designed to determine the minimum pixel size required in digital chest radiographic systems, we have studied 1 mm pixel digitized images on patients with early pulmonary sarcoidosis. The results of this preliminary study suggest that neither mild interstitial pulmonary infiltrates nor other abnormalities such as pneumothoraces may be detected reliably with 1 mm pixel digital images.

The routine measurement of exposures for a reference dimension patient in diagnostic radiology is regarded as an important part of an effective quality assurance program. The most frequent radiologic examination conducted in the United States is chest radiography. If manual techniques are used to conduct the exam, the procedure for measuring exposure to the reference patient is straight-forward. However, if automatic exposure controlled (AEC) techniques are used, a patient-equivalent chest phantom must be employed to reproducibly attenuate the x-ray beam. This is of particular importance if exposures are to be compared among AEC systems with different entrance x-ray spectra. Exposure monitoring is just part of the quality assurance story. Radiographic techniques, filtration, scatter reduction, film/screen use, and film processing performance (among other factors) must also be assessed, in order to effectively evaluate and modify these exposures so that they provide appropriate image quality. The first four factors are relatively easy to determine through measurement or documentation. Poor processor performance, potentially a major cause of abnormally high patient exposure, is more difficult to assess. We have designed, constructed and tested a Lucite/aluminum patient-equivalent attenuation chest phantom (LucAl) to use in the estimation of standard posteroanterior (22-23 cm) patient exposures for both manual and AEC chest systems. A sensitometric procedure that can be used to assess relative processor performance has also been developed. This paper describes these two procedures and their use in a six-State pilot study to monitor and evaluate exposure, technique and processor data in chest radiography. Results from approximately 200 chest systems will be summarized.

The role of the video camera in the typical fluoroscopic imaging chain is reviewed and measures of camera performance, including sensitivity, spatial and temporal resolution, and noise, are described. Finally, the characteristics of several pick-up tube types used in medical imaging are discussed.

The television camera comprises an important link in the imaging chain of digital subtraction angiography equipment. Various factors including spatial resolution, signal to noise ratio (SNR), progressive and interlaced read out, exposure utilization and camera lag are investigated. Requirements for the video camera for optimized DSA studies include sufficient bandpass to satisfy digitization matrix sizes, an 800:1 camera SNR, progressive read out of the camera target, and bias light to minimize build-up lag response.

This discussion describes a self-regulating, video hard copy camera system and compares its performance, for stability over time and temperature, to film processor sensitometry. Conclusions are made, regarding the means of identifying the source of image drift. Camera response curves are compared to display response and examples of raster line smoothing are discussed, to understand the cause of poor comparisons between the hard copy results and the video display.

The first fifty years of magnetic recording centered around one basic format. It was, and is, called longitudinal recording. The medium it-self progressed from wire, first used by Poulsen, to steel tape, to magnetic-coated paper tape, to the magnetic-coated acetate tapes commonly in use in the early 1950's. Quality of reproduction steadily improved over the years as well, and by 1952, magnetic tape recording was becoming the standard for audio mastering, broadcasting, and instrumentation (data) recordings.

Video systems are playing an ever increasing and more important role in diagnostic imaging. However, very little emphasis is being placed on the proper specification, evaluation, set-up or quality control of such systems. Even more disconcerting is the fact that the majority of video systems in use in diagnostic imaging do not conform to video standards making interchangeability of components and the evaluation of video signals difficult, if not impossible. The basics of video imaging systems will be discussed and video standards will be outlined. The basics of quality control of video components and display devices will be presented, while pointing out problems and pitfalls associated with testing these systems. In addition, some suggested acceptance limits will be provided.

The cathode ray tube, with raster scan display format, is commonly applied to a variety of medical instruments. In most cases the end product is to be a film transparency duplicating that obtained from older, conventional x-ray equipment. An attempt has been made to force the cathode ray tube to mimic imagery which would be produced by the phosphor in an "x-ray machine". The transition has not been without its difficulties.

Since the invention of x-rays by Wilhelm Conrad Roentgen, rapid advances have been made in the radiological detection of body abnormalities. This was very evident in the 1960's and 70's when the marriage of computers to radiology gave birth to a new generation of imaging modalities such as computerized tomography, ultrasound, digital radiographic imaging, nuclear medicine, and nuclear magnetic resonance. Many of these devices employ digital computer techniques for signal manipulation, and the resultant analog diagnostic images are displayed on television monitors for viewing and on imaging cathode-ray tubes for a photographic hard copy.

The broad speed classes into which film-screen receptor systems are now commonly grouped create considerable confusion and uncertainty in relating exposure and dose to properly exposed films. Practical speeds become even less certain by relating the whole speed scale to a reference system which is not uniquely defined. A system to overcome these difficulties is proposed, in the form of a scale of absolute speeds, based on a recent ANSI test method, and grouping resultant speeds into narrower classes. Data are presented which relate this recommended speed class width to allowable density ranges and to typical distributions of product speed.

An orthochromatic x-ray film made with tabular silver halide grains has a significantly higher MTF when exposed with green-emitting intensifying screens than do conventional films with similar sensitometric properties. The primary reason for the improved MTF is a decrease in the amount of crossover exposure, i.e., exposure by light that has crossed the support one or more times. Two well-established sensitometric procedures for measuring crossover have been compared. One produces results accurate enough for calculations of MTF relationships. Calculated MTF relationships for tabulargrain and conventional films are compared with measured values.

Measurements of the output signal-to-noise ratio for several screen-film systems are made in terms of the noise equivalent number of recorded quanta (NEQ). Described in this way, the signal-to-noise ratio is shown to be a strong function of both exposure and spatial frequency, with significant changes from system to system.

Anti-crossover films are now available for use with some rare earth intensifying screens. Two such films (Kodak T-Mat G and 3M XUD) were compared to a more conventional emulsion film (Kodak OG) under clinical conditions. Both anti-crossover films exhibited an observable improvement in sharpness, slight increase in visualized mottle and contrast. Speed differences were also observed. Some problems were encountered, but the consensus of most radiologist observers was an improvement in visualization of radiographic information because of this increased sharpness.

A test object for evaluating the performance of medical x-ray television systems has been developed. The test object consists of an array of six steel wires of varying diameters (range 0.022" to 0.005") arranged at 30 degree intervals like spokes in a wheel, sandwiched between two circular pieces of 1/8"-thick acrylic. This test object is viewed during fluoroscopy through various thicknesses of water (6"-10") to simulate small to large patients. It can be viewed while stationary or while rotating at 7.5 - 30 RPM. Standards of performance will be given.

Clinical medicine needs a convenient, relatively noninvasive technique for the measurement of regional arterial blood flow. Measurement of flow would be extremely useful in evaluating the improvement in flow following vascular surgery or angioplasty. A number of techniques based on film or video densitometry have been used.(3,4) More recently digital subtraction angiography has been used for the measurement of blood flow.(1,2,5) These techniques basically obtain the blood flow by measuring the time a bolus of contrast material takes to move between two points of an artery. By knowing the distance between the points, the average diameter of the vessel, and the bolus transit time the blood flow (cc/min) can be calculated.

Standard angiography demonstrates the anatomy of arterial occlusive disease but not its physiological signficance. Using intravenous digital subtraction angiography (DSA), we investigated transit-time videodensitometric techniques in measuring femoral arterial flows in dogs. These methods have been successfully applied to intraarterial DSA but not to intravenous DSA. Eight 20 kg dogs were instrumented with an electromagnetic flow probe and a balloon occluder above an imaged segment of femoral artery. 20 cc of Renografin 76 was power injected at 15 cc/sec into the right atrium. Flow in the femoral artery was varied by partial balloon occlusion or peripheral dilatation following induced ischemia resulting in 51 flow measurements varying from 15 to 270 cc/min. Three different transit-time techniques were studied: crosscorrelation, mean square error, and two leading edge methods. Correlation between videodensitometry and flowmeter measurements using these different techniques ranged from 0.78 to 0.88 with a mean square error of 29 to 37 cc/min. Blood flow information using several different transit-time techniques can be obtained with intravenous DSA.

Digital subtraction angiography of the breast may permit the differentiation of benign and malignant breast lesions. We have developed specific techniques for performing DSAB. The patient is examined in an oblique prone position with the involved breast in an immobilization device of our own design. The immobilization device adapts to our angiographic patient table and provides a water bolus with slight compression. The central ray of the x-ray beam is positioned for a lateral view of the breast, similar to the lateral view obtained in a mammogram. Iodinated contrast is injected from a catheter position in the superior vena cava. A kilovoltage of 50 kVp is employed which produces a near optimal signal to noise ratio for iodine contrast. The iodine signal to noise ratio characteristics of breast DSA have been modeled using a computer program which estimates the x-ray spectrum, filtration effects(tube, tissue, iodine, and grid), and image intensifier energy absorption. The energy absorbed in the input phosphor of the image intensifier is determined using a Monte Carlo radiation transport technique. Images are acquired in a 512 x 512 x 10 matrix with a 9" image intensifier using a geometric magnification of approximately 2. Typically, 10 mAs per exposure is required. A maximum of 40 exposures are made in three phases totalling 5 minutes. The average absorbed dose to the breast for a single exposure is 48 millirads (6 cm thickness) as determined by a Monte Carlo radiation transport computation of energy absorbed in breast tissue.

There is a number of Digital Subtraction Angiography (DSA) phantoms available in the commercial market that are designed for the evaluation of DSA imaging systems. This paper focuses on the performance evaluation features provided with these phantoms as seen from daily quality assurance point of view, and for possible usage in the acceptance testing of DSA imaging systems.

Two basic approaches to digital radiography have been thoroughly investigated in the last decade. In one, the traditional video chain is modified, and the output of the camera is digitized and analyzed by a computer to produce an image of the area of the body being examined. The other approach uses a linear detector array to generate a sequential series of line images as the array is scanned across the area being observed; the line images are then combined by a computer to form a two-dimensional image. We are investigating the characteristics of a third method, a prototype digital radiography system in which two-dimensional diode arrays (CCD) are fiberoptically coupled directly to either a scintillating fiberoptic plate or to a fluorescent screen. In this paper we describe the- concepts and design configuration of this approach, as well as preliminary results from several phantom studies. Our results indicate that high resolution, high signal-to-noise ratio imaging can be attained with this method.

The anticipated benefits of digitized projection radiographic data to medical imaging has generated much interest and activity. Numerous methods for capturing the radiographic image data prior to digitization are being evaluated. In this paper the excellent performance of the film-screen system as a receptor for projection radiographic data is discussed. An experimental system for obtaining high quality digital radiographic data by laser scanning radiographic films is described. This system is being used to evaluate the clinical utility of various digital image processing algorithms. Future plans include an investigation of quantitative analysis of projection radiographic data. Digital data obtained by film scanning can be used with digital image archiving and communications systems.

Screen-film mammography is growing rapidly. The impor-tance of using dedicated x-ray units with appropriate beam quality and breast compression for screen-film mammography is emphasized. The present status of grids for mammography is presented. Mammographic screen-film characteristics, in terms of contrast, light diffusion, and noise, are reviewed. Radiation dose in mammography is discussed in terms of measurement, calculation, and theoretical risk.

Digital imaging in one form or another is being used for virtually every aspect of medical imaging. Many are predicting that the diagnostic department of the future will be entirely digital. However, upon careful consideration, many questions remain to be answered concerning the monumental quantities of digital data which must be acquired, processed, transmitted, displayed, stored, and recalled. In order to be acceptable to the clinician the digital systems of the future must be able to produce images with image quality which is at least as good as that presently available, at costs similar to those incurred in today's imaging environment, and with the same ease of use as the clinician experiences today. If future digital systems only meet these goals without providing additional significant benefits then one must wonder if digital imaging will ever be adopted as the sole means of image acquisition, display, and storage. This paper attempts to answer the question--is the total digital medical imaging department of the future a dream or a nightmare?

A 512 X 512 digital store has replaced a video disc as a storage mechanism during pulsed fluoroscopy. This system, storing 1 TV field from a 1023 line signal following each pulse, is much more stable than the analog disc and can also reduce the fluoroscopic dose by 75%. This stability now makes the concept of pulsed fluoroscopy acceptable from the clinical point of view. Furthermore, the stored images on this matrix can resolve to 1.5 line pairs per mm, which produces useful permanent hard copy. This represents a further extension of clinical videofluorography, already developed by the authors. Its implementation for storage of gastro-intestinal examinations will be discussed, with assessment of the relationship of x-ray dose to image quality.

A digital radiography system using self scanning linear diode arrays is being developed for improved diagnosis at reduced radiation dose. Our technique is based on the use of solid state sensors with 1024 diodes per inch and with very high dynamic range. The slit geometry of our method results in image improvement and dose reduction by efficiently rejecting scattered x-rays in the patient. In our present configuration the images have a field of view of six inches by six inches or 6 inches by 12 inches and are digitized to 1024 x 1024 pixels with 12 bits. This digital system differs from the conventional digital radiography in that no image intensifier TV fluoroscopy chain is required. Preliminary clinical studies have demonstrated the high detail of our system at low radiation levels. In dog studies the system has clearly visualized very small coronary arteries following aortic root injection of contrast material. Even with intravenous injections some of the coronary arteries could be seen.

The SMPTE test pattern has proven to be an effective tool for calibrating and monitoring the image display devices of a magnetic resonance (MR) imaging system. Linearity and size adjustments of video displays are particulary important because of the proximity of magnetic fields. The 5% and 95% intensity levels of the test pattern are extremely useful for adjusting the grayscale of both video displays and multiformat hardcopy devices. An appropriate sequence of operations for adjusting and monitoring image display devices is recommended.

Respiratory motion has been observed to produce artifacts in MR images of the abdomen and thorax. In order to experimentally characterize this artifact, a phantom was moved cyclically in the imaging field in order to investigate the effects of 1. amplitude of motion, 2. speed of motion, and 3. direction of motion. All images were obtained with a 0.30 Tesla magnet using a spin-echo sequence and two-dimensional Fourier transform, 2DFT, reconstruction. In 2DFT imaging a steady frequency encoding gradient is applied during MR signal acquisition which is preceded by a sequence of varying phase encoding gradients which are perpendicular to the frequency encoding direction. The artifact due to periodic motion consists of repeated copies of the edge of the object which are displaced from the original object in the phase-encoded direction. The distance between the actual object and the first echo is inversely proportional to the number of MR signal acquisitions per motion cycle. The clarity of both the artifact and the object is reduced as the amplitude of motion is increased. Simulations of motion patterns confirm experimental results. Respiratory gating has been applied in both abdominal and cardiac studies. These studies resulted in improved image clarity and elimination of repeated copies of organ edges in the reconstructed image.

In magnetic resonance (MR) imaging, improved signal to noise ratio (S/N) can be achieved with higher magnetic field strength. Such improvement, however, will be limited eventually by the absorption of the radiofrequency (rf) within the patient. Radiofrequency propagation into an infinite cylinder of physiological dielectric properties has been calculated at frequencies from 1 to 100 MHz for a concentric single loop antenna. The amplitude of the rf wave at the center of the phantom can be either enhanced or attenuated depending on the dielectric properties of the phantom. Measurements performed at 25 MHz in a 28-cm diameter cylindrical phantom of 154 mM saline show an attenuation of 16% and a phase shift of 57° which agree well with the calculations for this medium.

Traditionally, multiple format recording emulsions for medical video imaging have utilized a film (transparent) base. The major reason for this is probably because the film and camera manufacturers felt the diagnostician is accustomed to viewing x-ray images on a film base and would prefer to view video images that way also. Because of the need to keep radiation exposure to patients at a minimum and the fact that photographic emulsions are generally very inefficient in utilizing x-ray radiation, a film base was the logical requirement for direct x-ray imaging as it enabled the image to be recorded by two emulsions rather than one. The transparent base thus allows viewing a photograph which is the result of the additive effect of the two emulsions. The use of transparent base imposed specific requirements that necessitated the development of a whole complex of equipment designed for the particular use of film such as the processing machines, their chemical solutions, and the famous viewbox and alternators that characterize the radiology departments of today.

The effects of CT display window width and lesion size on observer performance were studied by measurements of contrast-detail "thresholds" and lesion detectability (using ROC analysis). Both methods of measuring observers' perceptual performance showed that lesions became less visible as the width of the display window increased, and that the magnitude of this effect was greater for larger (lower contrast) lesions than for smaller ( higher contrast) lesions. Manipulations of the CT display window did not affect the information content of the physical image, as measured by actual calculations of signal-to-noise ratio for a "lesion-matched" filter. Therefore, the effects of display window on observer performance were attributed to psychophysical properties of the human visual system.

The recent awareness of higher radiation levels associated with some of the latest CT scanners points to the need for dissemination of meaningful quantitative dose data about these valuable imaging units. Examples of such data, from radiation protection surveys, will be presented. Such information should afford CT users a basis for proper utilization of this modality within current radiation protection guidelines.