Over the past several years development and refinement of medical imaging devices has occured at an exponential rate. Computed whole body tomography, grey scale ultrasonography, real time ultrasonography, radionuclide imaging with new radiopharmaceuticals, and digitized radiography are all currently available diagnostic imaging modalities. While each of these imaging techniques has certain unique advantages, the rapid development, introduction and promotion of these new diagnostic imaging modalities has infact created a crisis of plenty. Clinicians are faced with the dilemna of choosing from among the many techniques available; the most appropriate modality for a particular clinical situation often before adequate understanding of the capabilities and limitations of each imaging modality has been acquired.

The current state of medical diagnostic ultrasound imaging is analyzed including discussions of present limitations and possible solutions of these problems. Ultrasound is unique in that tomographic images are formed from mechanical energy reflected from tissue interfaces of differing acoustic impedance. The low propagation velocities of ultrasound (1540 m/sec) coupled with the range of diagnostic frequencies (1-30 MHz) result in excellent image resolution (2mm x 2mm) and permit electronic signal processing to achieve portable real time dynamically focused scanning systems at relatively low cost. Diagnostic ultrasound is applied in obstetrics, cardiology, abdominal imaging and ophthalmology but it versatility is limited by the physical characteristics of tissue. The relative impenetrability of bone and tissue containing air hinder its use in the head, lung and intestines. Tissue inhomogeneities throughout the entire body make it impossible to achieve diffraction limited resolution, and the slow propagation velocities of sound place an upper limit on the image data acquisition rate. Research is ongoing to improve resolution by adaptive imaging techniques, and parallel processing will hopefully increase data acquisition. These improvements depend heavily on advances in electronics technology.

The revolutionary impact of computerized tomography has caused radiologists and imaging scientists to re-examine the utility of conventional projection radiography. in addition, the improved x-ray technology which has accompanied computerized tomography shows great promise in being applied to new versions of projection radiography with dramatic improvements in performance. Projection radiography has a distinct advantage in that an entire volume of interest is presented in a single image. In CT this same volume requires an array of cross-sectional images. Certain structures, such as blood vessels, are much more readily visualized in the projection mode. It would be very cumbersome to study the stenosis of a vessel by viewing successive crosssectional images. It would indeed be fortuitous if the vessel remained in a single CT section for any considerable extent. Projection radiography, however, measures the line integral of the attenuation coefficient rather than the attenuation coefficient itself as is done in CT. This property of CT has enabled the use of the very powerful display technique of windowing where a particular small range of attenuation coefficients are observed. in this presentation, a subtle change can be studied where the visualization is limited solely by the counting statistics and dose. In general, this windowing property has not been available in projection radiography so that radiographic images have tended to be limited by contrast, rather than signal-to-noise ratio. This is an undesirable situation which must be remedied if projection radiography is to significantly improve. Two important factors are required for improved projection radiography, improved detection and image processing. The existing detection systems have relatively low quantum efficiency, significant scatter and poor dynamic range. Recently, projection systems have been used with CT detection systems, such as the GE Scout View system. These essentially provide all of the desired properties with somewhat reduced resolution. In addition, a variety of digital image processing activities have been initiated including various subtraction modes and spatial filtering which provide visualization of structures which were otherwise contrast limited.

Nuclear imaging is a powerful diagnostic modality which obtains images from the intact body through the administration of miniscule amounts of physiologically specific radioactive agents (Table I). Typical radiation dosages do not exceed one rad to the critical organ, and the small molar concentrations of the pharmaceutical used mitigates carrier effects and poses no recognized hazards due to toxicity. With exception of imaging of the thryoid by fluorescent excitation of iodine, nuclear images reflect the very recent history of distribution of a physiologically specific agent or its selective uptake.

Computed tomography's exceptional contrast sensitivity, which can show local changes in absorption coefficient of 0.5% of the absorption coefficient of water anywhere in the cross-section, was truly revolutionary compared to conventional radiography's ability to distinguish perhaps a 2% change in absorption in the total transmission path through the object. This contrast sensitivity combined with 3-D localization has given CT a preeminent position in neuroradiology and has lead to the rapid development of CT for applications in the rest of the body. To date, almost all of these applications have involved static or steady state scans. One current extension of CT is its use in transient studies, for example to follow a bolus of radiopaque in its passage through an organ such as the heart. This technique has been used to demonstrate regional blood flow abnormalities in animal experiments and coronary bypass patency in patients. The performance of CT scanners is well understood and major performance characteristics can be quantitatively calculated or predicted. These include both high contrast resolution and low contrast detectability, dose, and even some types of artifacts. For conventional static scanning the most important questions are not technical, but relate to the clinical utility and cost effectiveness of various possible designs. For designs aimed at transient studies, x-ray source brightness is probably the most significant technical limitation. Most CT scanners are close to being source brightness limited, with room for only modest improvement before system trade-offs are required. The resulting system constraints combined with CT's high visibility in society's effort to control health care costs suggest that future CT developments are likely to evolve with considerably clinical interaction and guidance.

O. T. von Ramm: I'd like to ask a question of Dr. Moss. From the point of view of a clinician, what would you like the ideal imaging system to do?

Although water delay B-scanners were first developed in the early 1950's, they proved to be quite inflexible and awkward for the patient, who was actually submerged in a water tank. Since the mid 1960's water offset ultrasonic techniques have been refined at the Ultrasonics Institute of the Australian Government. National Acoustics Laboratories. There, the efforts of Dr. George Kossoff have resulted in the production of a practical B-scanning system (1,2). It is called the U.I. (Ultrasonics Institute) Octoson.

This paper discusses the development of a new phased array sector scanner at Picker Corporation. Among its features are 256 sector lines in a 90° sector with a fixed transmit focus and a dynamic receive focus. Several sector angles are available. Charge-coupled devices (CCD's) are used as delay elements under control of a microprocessor and PROM memories. In the past dynamically focussed systems have used digitally switchable LC delay lines or CCD's driven by exponentially swept VCO's to cause the receive focus move with the receding acoustic wavefront. The present system uses a digitally controlled feedback stabilized frequency generator to establish the clocking frequencies for the CCD's. The numerical control sequence for this generator is designed to minimize the delay error accrued by the signal as it traverses the CCD. The quality of the beam formation is demonstrated by results of tank tests.

The number of bits required for storing the information contained in an x-ray image depends primarily on the x-ray exposure at the image plane and on the desired spatial resolution. This dependence has been determined from calculations of the maximum attainable signal-to-noise ratios for exposures and resolutions respectively in the regions from 10^-2 to 102 milliroentgens and 10^-4 to 1 cm2. The results show that for the different image sizes and exposures used in diagnostic radiology, the number of bits required for the storage of these images extends from approximately 10⁵ to 10⁸.

Six sources of inefficiency in the radiologic imaging process are outlined. Each source is subject to further improvement by a factor just less than two, leading to a possible overall improvement, or radiation reduction, of a factor 20 to 40. Practical technological considerations suggest that a six- to tenfold reduction is realizable over the next decade.

The first practical CT scanner was design0 for the head and it took 4 1/2 minutes to collect data for a pair of cross sections'. Today's scanners take less than 5 seconds to collect the data and are designed to produce cross-sectional images of any part of the body. However, the heart has remained a challenge for CT because of the problems associated with its motion2. To avoid motion problems, much of the early cardiac CT work has been in vitro studies or has examined arrested canine hearts3. Also, gating has been used to produce "stop action" or strobe type images of in vivo beating hearts4. Since gating uses data collected over many heart beats, it can only be used to study periodic phenomena in the steady state. Thus, gating cannot be used in a very important diagnostic area, the study of flow or perfusion using contrast medium or x-ray absorbing dye injected into the blood stream. One group has started construction of a very fast scanner which is designed to scan the entire heart every 10 milliseconds5. This design has had to sacrifice appreciable signal-to-noise ratio (contrast sensitivity) for scan speed, a good trade off for imaging anatomy, particularly in systole when the heart is moving most rapidly, but not necessarily the best choice for transient studies such as the perfusion of contrast medium in the heart muscle, where the signal-to-noise ratio is of paramount importance. Our approach has been to explore the potential of present scanner hardware for transit and cardiac studies. We find, in agreement with most previous work, that the x-ray absorption coefficient of blood and heart muscle are so close to being equal that CT scans without the use of at least small quantities of radiopaque contrast medium are of little value. However, we also find that with the contrast between muscle and blood enhanced by the use of a contrast medium, motion problems are not as severe as had been anticipated and that many useful studies, both static and transient, can be performed with a 5 second scanner. Some examples described here are: assessment of the damaged (infarcted) region following a heart attack, visualization and measurement of abnormal wall motion, regional perfusion in the heart muscle (myocardium) and the condition of coronary bypass grafts.

This paper represents the third presentation to the SPIE that has dealt with our research activities in photoelectronic radiology and intravenous angiography. The earlier papers have been published in Volumes 1271 and 1642 of the Proceedings, and cover the evolution of the technical facilities and associated radiographic images obtained with dogs. We have now reached a level of competence suitable for the examination of patients, and we present examples of the images obtained to date. Although the research activity reported here began with the task of non-invasive imaging of arteries for the early detection of atherosclerosis, it is becoming increasingly clear that the method is applicable to all areas of the body to which angiography is being currently applied. Furthermore, with the advantages of this technique, new applications are being developed. Thus, we are now beginning examination of animals and patients to determine the utility of intravenous (non-invasive) angiography for examination of the head, coronary arteries, heart, lungs, kidneys, and extremities, and look forward with considerable optimism to the results.

The detection stage of any medical imaging system can be quantified rigorously using information theory concepts developed by Shannon. The motivation for this approach is outlined and an example of an application to computed tomography systems is given. Relationships to traditional noise/modulation transfer function analyses and to noise equivalent quanta (NEW and detective quantum efficiency (DQE) concepts are explored.

The feasibility of an x-ray imaging device based on photoconductive control of an electrophoretic display has been explored. Contrast ratios of 18 to 1 and resolutions of 8 line pairs/mm have been measured. The radiation to reflectance transfer characteristic is non-linear, with substantially less sensitivity to small exposures. Radiation exposures of 2.4.mR have produced satisfactory x-ray images. Excellent quality images have been obtained but cell stability and linearity need to be improved.

Scatter radiation in diagnostic radiology is recognized as an agent that adds noise to the imaging system and increases patient dose. A prototype computed radiography (CR) system by General Electric uses a highly collimated, 1.5 mm FWHM, pulsed fan-beam of x-rays, good geometry and well collimated xenon gas detector array to virtually eliminate scatter radiation resulting in high contrast sensitivity and low patient dose. The detectors with their high signal-to-noise ratio, wide dynamic range and large x-ray quantum efficiency provide added benefit in terms of wide latitude and reduced dose to the patient. The CR system is found to use radiation more efficiently than film-screen-grid standard radiography systems for imaging low-contrast objects in parts of the body such as abdomen.

Region-of-interest tomography (ROIT) reconstructs a convex region within an object on a specified sampling grid with relatively small dose and higher resolution. In this paper, we introduce a new algorithm for ROIT. This algorithm is based on a variable sampling scheme in which the external region is sampled coarsely whereas a finer sampling is used for the region of interest. After the data is collected, the coarsely sampled data is interpolated at the sampling interval as the finely sampled data. The interpolated coarsely sampled data and the finely sampled data are assimilated and entered into a reconstruction of the ROI. We have obtained encouraging results in experimental computer simulations of our technique. These results are summarized in this paper.

A new approach to the reconstruction of images with limited data from rotary fan-beam computed tomography (CT) scanners is presented. The missing views are estimated by reflecting the data obtained at source positions in the range (l80±fan angle) away from the desired viewing angle. Due to asymmetry of the source/ detector geometry the "reflected" views may be offset from the desired views and additional corrections are required. Measured and "reflected" views are convolved and backprojected using the standard fan-beam geometry reconstruction algorithm. This method is used to reconstruct 1.8 second images from a rotary fan-beam scanner which rotates 360° in 3 seconds, providing improved temporal resolution for dynamic CT studies with no increase in noise at a constant radiation exposure. In addition, the method is applied to ECG-gated cardiac imaging, reducing the scanning interval required for generating systolic and diastolic images by a factor of 40%.

Rapid cardiovascular computed tomography (CT) requires image reconstruction from a limited set of projections. Deterministic methods using convolution back projection algorithms have not been suitable for these limited data cases. An alternative approach has been formulated to find the minimum mean squared error image estimate using a Kalman filter with polar pixels for two-dimensional reconstructions of both simulated phantoms and real objects from data obtained on a rotate only CT scanner. Computation time was minimized by limiting the number of pixels to 120 and using a rotationally symmetric (polar) pixel structure. The Kalman filter was compared with Algebraic Reconstruction Technique (ART) for full view, limited view, and missing view measurement sets. The Kalman filter performed with consistently lower mean squared error than ART for both real and simulated data and rapidly converged to the theoretical limit of resolution. Performance of the Kalman filter was optimized only if the system noise (error) was adequately characterized. When real objects were scanned it was necessary to include the measurement errors introduced by finite pixel width and finite beam width in addition to Poisson noise to achieve optimality. The use of polar rather than rectangular pixels provided a reduction in computation and storage requirements for the Kalman filter. This study demonstrates the potential utility of Kalman filtering methods using polar pixels for limited data CT image reconstruction. Common to the methods used to create cross-sectional images through various anatomic structures using x-rays, gamma rays or ultrasound is the mathematical reconstruction of a two-dimensional image using a series of projection measurements represented by line integrals. While considerable effort has been expended in the theory of image reconstruction from line integral measurements there are still many applications of cross-sectional imaging where the mathematical methods of computation are inadequate. Frequently used reconstruction algorithms assume that (1) projection data are symmetrically arranged; and (2) there are no sources of noise (i.e., all measurements are exact). In reality neither assumption is true, and for certain applications, e.g., rapid scanning of the heart, minimizing patient dose, etc., these assumptions are so flagrantly violated that these algorithms produce images with severe artifacts. In this paper we demonstrate the use of a new algorithm2 for image reconstruction which yields in the least squared sense optimal solutions in the presence of noisy and asymmetric measurement sets. Results indicate that this method could allow more accurate cross sectional imaging than is possible with conventional methods.

A new CT imaging technique is described. It does not employ x rays; instead, weak electrical currents are passed through the patient to map out the electrical properties of the tissues. It is expected that the images will ultimately have sufficient precision and resolution to be of clinical value.

Nuclear magnetic resonance (NMR) imaging has already attracted a great deal of attention because it is non-hazardous and because the intrinsic NMR contrast appears to be considerably higher than x-ray contrast. On the other hand, the in-vivo NMR characteristics of normal tissues are not well understood, and no reliable data exist on the potential for differentiating normal from abnormal tissues, or benign from malignant lesions.

A new x-ray perspective and tomographic imaging technique which employs the field backscattered by an object has been demonstrated to have interesting medical applications. The resulting radiographs appear as if the objects' internal features were interrogated by an "x-ray eye." Uncollimated detectors are placed on the same side of the object as the incident radiation. A flying spot technique creates a backscatter image. Spatial resolution of the system is dictated by the diameter of an x-ray beam which raster scans the object under study. One millimeter resolution at penetration depths of several centimeters of tissue equivalent material is demonstrated with the use of short pixel dwell times to achieve radiation dose levels of the order of 10 mR. A system which generates real time x-ray backscatter images was constructed and used to examine biological and test objects. Two modes of image formation have been employed: 1. The unusual perspective of an x-ray backproj ection of internal features. 2. The differentiation of different peak voltage scans to obtain longitudinal tomographic images. The system is sufficiently practical with several medical x-ray sources that medical diagnostic radiography could immediately derive benefit from the changed point-of-view.

A novel method has been developed to evaluate the noise of x-ray image intensifiers. Fast electronics and a fast photomultiplier tube (PMT) are optically coupled to the output of an x-ray image intensifier. The light emission induced in the intensifier by the absorption of x-ray photons is measured by counting single PMT photoelectrons. The fluctuation in the number of counted PMT photoelectrons per absorbed monochromatic x-ray photon is a measure of the noise of the x-ray image intensifier. It is characterized by an efficiency factor called DQEscin whose values have been obtained from PMT count distributions for five x-ray energies. Experimental results reveal that the output signal-to-noise ratio of the x-ray image intensifier under study is reduced by no more than 10% as expected from the efficiency factor DQEscin.

A brief review of the development of computer analysis of chest X-rays at the University of Missouri - Columbia is given. Chest analysis by computer has developed from cardiac silhouette shape and size analysis to lung field analysis in the form of image texture algorithms. The overriding problem is the variability in the normal human anatomy, especially as it is presented in radiographs. A major effort has been made to produce and verify valid data bases of substantial size. The diagnostic effectiveness of digital image analysis algorithms has been the utmost concern. Success has been achieved on a small scale relative to the total area of automated chest X-ray analysis.

A simple technique has been developed to simultaneously display regions of a CT image which have large differences in CT numbers, such as lung and soft tissue. The CT image is considered to be the sum of two unimodal distributions of CT numbers and the CT numbers associated with one region are mapped into the other using a simple linear transformation. The significance of this technique is that it permits the entire CT image to be visualized with optimum contrast either on the CT display monitor or on a single photograph. Examples of a body section and a head section are presented.

Providing the clinician with an accurate visual representation of digital nuclear medicine data requires 1) interpolation to fill in the intensity field between grid points, 2) correction for grayscale nonlinearities inherent in the display and film, and 3) sufficiently fine graylevel resolution to avoid generating artificial contours. Results from preliminary experiments using a precision computer display/film system have been encouraging, indicating improved image interpretation is frequently possible compared with both conventional analog scintigrams and commonly available computer displays, A wider range of count rate data was visible in the digital images giving better identification of low count rate areas, display artifacts due to regularly spaced data samples were eliminated as were contour artifacts caused by too few graylevels, and the relevant anatomy and or pathology was frequently demonstrated with greater clarity. Clinical examples will be presented which illustrate the benefits to be gained by using these techniques.

Object location in computed tomography images is a preliminary step required for automated measurements which may be useful in many diagnostic procedures. Most object location, image processing techniques are either globally based such as histogram segmentation or locally based such as edge detection. The method described in this paper uses both local and global information for object location. The technique has been applied to the location of suspected tumors in CT lung and brain images. Sorting and merging steps are required for eliminating noise regions but all suspected tumor regions have been located. Measurements such as boundary roughness or density statistics may also be made on the objects and used to identify suspicious regions for further study by the radiologists. Algorithms for chain-encoding the object boundaries and locating the vertices on the boundaries is also presented and compared. These methods are useful for shape analysis of the regions. The significance of this technique is that it demonstrates important additional capability which could be added to the software libraries of most CT systems.

A pipeline approach to processing of digitized picture data using multiple mini computers and an array processor is presented. Picture size can be up to 512 by 512 pixels. Implementing of many heuristic algorithms such as edge detection, edge enhancing, convolution/ correlation of template, peak detection, fast fourier transforms, filtering, summing, two pictures, registration, histogram equalization, thresholding and object counting, is determining and application is made to a nerve fiber counting project. The goals of the medical project are to provide data to be used in determing optimal surgical repair of injured or severed nerve and time scale as well as a percent of recovery of function following surgical repair. A simple operating system is described to invoke specific routines in the needed order in designated processors. Three approaches are discussed to the problem of image enhancement, pattern recognition and display of a picture consisting of multiple scenes or an object which is captured in the form of multiple "slices" through the object.

Over the past five years, our group at the Arizona Health Sciences Center has been developing a system for photoelectronic radiology. One of the projects in which we are involved is intravenous angiography, which Dr. Paul Capp reported on in Session 2 of Recent and Future Developments in Medical Imaging II. The purpose of this paper is to show some of the procedures of manipulation and measurements that have been developed to obtain better subtracted images.

Computed radiography (CR) is a recent development in diagnostic radiology which yields digital radiographs. Digital image enhancement of CR images in the form of smoothing the noise and enhancing the edges of anatomic boundaries has been used as a means to aid the physician in extracting clinical information from the radiograph. Details of the smoothing and edge enhancing function are discussed along with potential diagnostic applications.

The problems of intelligent image processing by computer, especially the processing of medical images like computed tomography scans, is examined in light of current image segmentation techniques. It is concluded that part of the problem lies in the lack of knowledge about how to guide low-leveiprocesses from higher level goals. An iterative boundary-finding scheme is presented which may aid in this guidance, and results from using specific criteria in the general framework to locate kidneys in abdominal computed tomography scans are presented and discussed. The problem of complex object localization in images is discussed, and some avenues for further research are indicated.

The perception of edges on computed tomographic (CT) scans appears easy but in fact is difficult. Such perception is important because it is necessary to make quantitative determinations. Diagnosis of such entities as spinal stenosis (narrowing of the spinal canal with encroachment on spinal cord and nerve roots) hinges upon an accurate knowledge of cross-sectional areas.

Robert Hooke conjectured about fluid circulation in plants as well as in animals in Micrographia in a passage that is equally important as a commentary on the dependence, not of technology on science, but of science on technology: It seems very probable that Nature has ... very many appropriated instruments and contrivances, whereby to bring her designs and end to pass, which 'tis not improbable but that some diligent observer, if helped with better Microscopes, may in time detect. This paper, written in the form of a scientific poem, reviews the current and near-future state-of-the-art of automated intelligent microscopes based on computer science and technology. The basic concepts of computer intelligence for cytology and histology are presented and elaborated. Limitations of commercial devices and research proto-types are examined (Dx), and remedies are suggested (Rx). The course of action pro-posed and being undertaken constitutes an original contribution toward advancing the state-of-the-science, in the hope of advancing the state-of-the-art of medicine. With rapid, contemporary advances in both science and technology, it may now be appropriate to modify Hooke's passage: It seems very probable that Nature has ... very many appropriated instruments and contrivances, whereby to bring her designs and end to pass, which 'tis not improbable but that some diligent observer, if helped with Intelligent Microscopes, may in time detect.

The volume of a polyhedron, which approximates an anatomical feature or lesion is calculated by means of an algorithm which is evoked by the Gauss' Theorem. Such polyhedra are useful in reconstructing surfaces of anatomical features and lesions from data obtained from CT or ultrasound scans. The volume estimation algorithm is presented. A proof of the algorithm is provided together with applications.

Boundary detection in conventional nuclear medicine scintigrams is often difficult for several reasons. First, scintigrams generally have a low signal-to-noise ratio. Second, edge structures are poorly defined because of the low resolution of gamma ray cameras; and finally, edge contrast is usually reduced by foreground and background activity. In this paper we report on heuristic approaches we have taken to solve these problems and to develop programs for the display of cardiac wall notion and for the automatic determination of left ventricular ejection fraction. Our approach to processing cardiac scintigrams entails several steps: smoothing, edge enhancement, thresholding, thinning and contour extraction. We discuss each of these steps in light of the goal of producing cardiac boundaries which are spatially and temporally smooth and continuous. Boundary detection results are presented for some selected clinical images.

This paper briefly reviews methods which have been employed to do image segmentation and indicates how texture analysis might be utilized to do this task. A method will be described for finding the unit cell size of a texture. The unit cell represents a tile which can be used to tile in the plane and generate the texture. It is the fundamental building block of repetitive textures.

An iterative method for the determination of the start and end of the tracer flow through a compartment was developed and used to create a digital image of the compartment of interest. This digital image was then used to select a region-of-interest corresponding to a selected compartment. Dynamic curves were corrected for background, dead time, and for possible overlapping of different physiological compartments. The shape of the corrected curves was then compared with expected behavior of the tracer in a compartment. On the basis of the physiological considerations, mathematical models are developed. Once the mathematical model is developed, the experimental data is used as suggested by the model. We applied these general considerations to the studies of the heart, lungs, kidneys and hepatobiliary system.

The method of contrast-detail-dose analysis offers a practical way of obtaining the information content per unit dose to the patient on an imaging system. The method is described and its assets are discussed.