A review of the processes by which ultrasound is generated through the interaction of ionizing radiation and irradiated materials: some possible applications.

The production of acoustic waves following the absorption of energy deposited by ionizing radiation, with a consequent production of localized thermal spikes has been confirmed by a number of papers published in the physics literature. This paper reviews the basic theory and presents most of the supporting experimental data. Some of the experimental methods used and the results obtained are summarized. In addition to the rather straightforward and routine use of acoustic phenomena produced by ionizing radiation for the detection and measurements of such radiation, there are some special applications that appear to be especially attractive for medical physics. Some of these are unique to ionizing radiation in that the amplitude of the ultrasound wave is proportional to the energy deposited in small volumes at localized sites of these interactions, while others derive from methodologies already in use with nonionizing radiations. The detection and measurement of this ultrasonic radiation could possibly lead to methods for the study of such fundamental phenomenon as track structure, precision localization of therapeutic treatment beams, and even the possible imaging of internal anatomic structures to provide on-line portal images.

Digital tomosynthesis: phantom and patient studies with a prototype unit.

A prototype digital linear tomographic system has been developed and used in patient and phantom studies. Phantom work demonstrated 0.6 line pairs/mm of spatial resolution, a slice thickness of 5 mm to 6 mm, range of reconstruction of 1 cm to 20 cm above tabletop, reconstruction time of 2 to 45 seconds per slice depending on matrix size, and a 280 mm field of view at the tabletop. Patient studies were conducted in three categories: renal imaging in 8 patients during intravenous urography (IVU); central airways imaging in 4 patients undergoing laser resection of endobronchial tumor; the knee of one volunteer. The digital tomograms from IVU studies were similar to the conventional film tomograms, and the only parenchymal abnormality, a renal cyst, was identified by both modalities. Digital tomograms of the mediastinum were superior to the chest radiographs in determining the patency of the central airways. With further development, digital tomosynthesis could offer greater spatial resolution, faster reconstructions, and less radiation than computed tomography (CT) or conventional tomography.

Patient exposure requirements for high contrast resolution in digital radiographic systems.

The contrast resolution of modern digital radiographic equipment is primarily limited by quantum mottle. The derivation of the contrast resolution of a system as a function of radiation exposure is presented. Calculations based on this derivation show that achieving 0.8% contrast resolution (128 meaningful gray levels) with 1-mm spatial resolution may require a patient exposure of 1.7 rad (1.7 cGy) per image of a thick body part (20 cm tissue). Acceptable clinical studies often can be obtained with only 5% contrast resolution (20 meaningful shades of gray), but even this level of contrast may require a patient exposure of nearly 90 mrad (0.9 mGy) per image. A comparison of three commonly used methods of reducing patient radiation exposure (pulsed imaging, temporal averaging, and recursive filtering) shows that, in theory, a pulsed system achieves the highest contrast resolution for the same total exposure.

Quantum noise in detectors.

The variance in the number of x-ray photons absorbed by a detector is derived from theoretical considerations. The distributions in the number of x-ray photons produced by the x-ray tube, transmitted through the patient, and absorbed by the detector are described. These distributions are compounded to yield a distribution in the number of x-ray photons absorbed by a detector. The standard deviation (quantum noise) in this distribution is very well approximated by the commonly used and experimentally verified quantity square root N where N is the average number of x-ray number of x-ray photons absorbed by the detector.

Normalization (division) versus subtraction of digitized images.

The propagation of noise in digitized images arrived at by algebraic manipulation of two images has been investigated. The algebraic manipulations investigated were logarithmic subtraction, linear subtraction, and division of the two images. The two images were identical except for the presence of a contrast-adding component in one of them. The two extreme cases of video-fluoroscope system behavior, gray level proportional to incident photon flux and gray level proportional to the product of thickness and linear attenuation coefficient, were investigated. The ratio of signal to noise was calculated for each of these cases and a figure of merit established for the ratio of signal to noise of the division process versus each of the subtraction processes for each of the two fluoroscopic system behaviors. In all cases, this figure of merit is greater than unity. Values of these figures of merit for iodine concentrations of 5, 20, and 300 mg/cm3 are given in the paper. The process of division appears to be advantageous as a signal-to-noise enhancer for all applications not requiring a quantitative relationship between image contrast and photon flux.

Validation of a computer-assisted method for estimating the number and volume of gallstones visualized by cholecystography.

It is likely that in the near future there will be widespread use of medicinal therapy to dissolve gallstones. The efficacy of medicinal therapy can best be determined by attempting to relate the total surface area of a collection of gallstones to the composition of bile in patients undergoing therapy. Surface area, in turn, can be directly related to gallstone size and number. In this study, involving 48 cholecystectomized patients, we have shown that standard cholecystography, together with a computer-assisted method of metrology, can effectively monitor the above parameters. Determinations of the standard deviation of 1) replicate readings (35.8%) and 2) averaged metrology estimates compared with actual stone volumes (42.9%), as well as correlation of actual stone volumes with averaged metrology estimates (r = 0.961), indicated the magnitude of assessed change in stone volume that would be necessary to accept a roentgenographic decrease or increase in stone size with 95% confidence. Even with the increased precision found in the computer-assisted method as described, to attain a 98% certainty of some volume change it was necessary to have metrology volume change of 50% or more. Actual stone counts were without significant error in 87.5% of the determinations.

Digitized longitudinal tomography.

An analog video longitudinal tomographic system has been converted to a digital format that reconstructs the tomographic section through the use of matrixes representing the gray scale values of the analog-recorded video images. This was done so that some simple preprocessing could be used to eliminate electronic noise and thereby increase the effective signal to noise ratio. Signal to noise was also increased by the elimination of the analog image displacement circuits (ramp generators). In addition to an improvement in the diagnostic quality (gray scale) of anatomic images, a capability to image contrast differentials as low as 1% has been achieved. A display system provides the ability to use windows having both variable central levels and widths.

Video techniques for x-ray imaging and data extraction from roentgenographic and fluoroscopic presentations.

This review is concerned with the use of video-fluoroscopic images in diagnostic radiology. It particularly emphasizes the potential of such images for performing quantitative procedures such as used in pulmonary and cardiac assessment, evaluation of contrast differentials, and tomographic (including CT) applications. This article also attempts to determine limitations, capabilities, and required developments for both the present and future hardware systems. Most importantly this review attempts to assess the present state of the art and to analyze the reasons for the limited application of such techniques in current radiology operations.

Electron backscattering.

Although electron backscatter has been a subject of investigation for many years, both theory and experimental data are sparse. This article examines the available data and shows that a large percentage of both the theoretical and experimental values fit an empirical relationship for a large range of electron energies. This relationship appears to be valid both for beta-ray spectra and monoenergic electrons. This relationship appears to be valid both for beta-ray spectra and monoenergetic electrons. The empirical relation describing the intensity of electron backscatter is also valid over the entire range of atomic elements. The relationship proposed is that the backscatter of electrons is proportional to the log (Z + 1). Effective atomic numbers of compound and mixtures arrived at using this relationship can be used to predict backscatter coefficients for these.

Enhancement of video images.

The enhancement of radiographic and fluoroscopic images using simple video analog techniques is described. In each instance, both the degree of enhancement and the features of the image to be enhanced are under the direct control of the radiologist. Noise is suppressed with a sharp cut-off, low-pass filter. Three types of analog circuits are discussed. One provides edge sharpening and contrast enhancement; one allows either black or white suppression, with expansion of the remaining shades of gray; and one provides an exponential response to selectable portions of the input signal.

Computerized system for the measurement of gallstone volume, in vivo.

Instrumentation and methodology for the determination of gallstone volume from standard roentgenographic views have been developed. Evaluation of size is made by using video viewing, an operator-set electronic window, intercept registers, and the programming of a PDP/8E for handling the data. The programming makes corrections for geometric factors arising from anatomical variations, corrections for variations in technique, roentgenographic magnification, and provides for the use of several roentgenographic views. The precision of the measuring (video) technique is approximately 0.5 mm as measured on the roentgenogram. Estimates and test data indicate that this system should be capable of an absolute accuracy within 10% of the true volume.

The capability of fluoroscopic systems for the production of computerized axial tomograms.

Previous work in our laboratories and at other institutions has shown that fluoroscopic images recorded on a video disc can be used successfully for producing computerized-axial-tomograms. The work described in this paper gives a quantitative analysis of the capabilities of such imaging systems, in conjunction with a particular method of data processing, for detecting and imaging changes in object absorptivity. Relations between the degree of contrast or absorptivity and object size required by this type of system can be inferred from the data.

Tumor localization and beam monitoring--electrofluorotomography.

Methodology and instrumentation are presented which are potentially capable of presenting fluoroscopically derived transverse axial body sections for use in radiation-oncology treatment planning and beam monitoring. These combine the methods of Takahashi for generating transverse axial tomograms with electronic radiography, electrofluorotomography, and a contouring program for extracting body and tumor contours in a digital format. The system will also be capable of assuring both initial and day-to-day beam alignment.

A computer program for the localization of small areas in roentgenological images.

A method and associated algorithm are presented which allow a simple and accurate determination to be made of the location of small symmetric areas presented in roentgenological images. The method utilizes an operator to visually spot object positions but eliminates the need for critical positoning accuracy on the operator's part. The rapidity of measurement allows results to be evaluated on-line. Parameters associated with the algorithm have been analyzed, and methods to facilitate an optimum choice for any particular experimental setup are presented.