A new modified bootstrap approach to sensitometry is presented which provides H and D curves that show almost exact agreement with those obtained using conventional methods. Two bootstrap techniques are described; both involve a combination of inverse-square and step-ped wedge modulation of the radiation field and provide intensity-scale sensitometric curves as appropriate for medical radiography. H and D curves obtained with these modified techniques are compared with those obtained for screen-film combinations using inverse-square sensitometry as well as with those obtained for direct x-ray film using time-scale sensitometry. The stepped-wedge of the Wisconsin X-Ray Test Cassette was used in the boot-strap approach since it provides sufficient exposure latitude to encompass the useful den-sity range of medical x-ray film. This approach makes radiographic sensitometry quick and convenient, allowing accurate characteristic curves to be obtained for any screen-film cassette using standard diagnostic equipment.

A systematic study was performed to investigate the accuracy and precision attained by four methods for determining the characteristic curve of radiographic screen/film systems. The four methods include: inverse square sensitometry, KVP adjusted bootstrap sensitometry step-wedge bootstrap sensitometry, and step wedge attenuation sensitometry. The inverse square method was used as the reference standard for accuracy. The extent of and the sources of the inaccurcies involved when using the alternative methods is discussed.

We have developed a new experimental method to determine the detective quantum efficiency (DQE) of any given x-ray screen for different x-ray energies. This method employs fast (10 ns) counting and coincidence techniques to directly measure the number, m, of photons emitted from the screen per absorbed x ray, and to determine both the statistical distribution and average value of m. These experimental data have been obtained for a calcium tungstate and a rare earth oxysulfide screen with quasi-monoenergetic x-ray beams in the region from approximately 15 to 70 keV. Examples of the experimental DQE values obtained from these measure-ments for pairs of the two types of screens respectively are 0.86 and 0.75 at an x-ray energy of 18 keV, and 0.20 and 0.41 at an x-ray energy of 69 keV. These results indicate a reversal in the relative capabilities of the screens to transfer image information such that there would be a preference for the calcium tungstate screen at 18 keV and the rare earth oxysulfide screen at 69 keV.

The major potential errors in the slit method measurement of the line spread function (LSF) and the modu-lation transfer function (MTF) of radiographic screen-film systems are reviewed. These errors are compared with the potential error introduced in the LSF due to an erroneous representation of the characteristic curve (sensitometric data). If a deviation in the sensitometric data results in a lower average gradient of the characteristic curve, the MTF will show higher resolution capabilities.

The effects of MTF, scatter radiation and display on the contrast-detail characteristics of medical film/screen radiography are examined. Theoretical relationships between these quantities are derived using signal-detection and effective aperture approaches. It is seen that scatter radiation significantly affects the perceptible information in the images according to the scatter degradation factor and is somewhat a function of subject contrast. It is shown that the dose efficiency of film/screen radiography can be improved by a factor of two by the elimination of scatter.

A method for the prediction of radiographic contrast rendered by screen-film systems is described. The observed invariance of H & D curve shape with x-ray energy is used to postulate the existence of a single characteristicyrve for a given system. Densities are then described from this curve in terms of an effective exposure S(E)0(E)dE, where S(E) and 0(E) are the energy dependent sensitivity of the system and the x-ray spectral fluence respectively. We used two experiments to demonstrate the validity of this approach for typical screen-film systems. Contrast predictions based on fs(E)o(E)dE were made for screen-film systems having different phosphor materials. Results, which show contrast advantages for calcium tungstate systems in the no scatter case are consistent with the calculations of others based on screen x-ray energy absorption.

Rotating aperture wheel (RAW) devices can be used in conventional tomography. This represents the possibility for the first application of efficient scanning slit anti-scatter techniques to this area of radiology characterized by inherently low contrast images. One coder wheel between the x-ray tube and patient and two scatter discriminator wheels between the patient and image receptor form sections of the RAW "projection cone" with the lines of radiation from the x-ray source forming the "flux pyramid". As long as the projection cone follows the motion of the x-ray flux pyramid (with the ratios of the distances between the x-ray source, RAWs, patient, and image receptor kept constant throughout the motion) any RAW pattern may be used. Simple relations are given which describe the geometric constraints for various tomographic motions. As in any application of scanning slit techniques, it would be possible to use the excellent scatter elimination capabilities of a RAW device either to improve image contrast or to reduce patient dose. Applications of rotating apertures to computed tomography (CT) are also possible. Multiple, simultaneous coronal or sagittal sections might be obtained with incomplete sets of projections using a single RAW system or with complete sets of data using two orthogonal RAW systems. A new device involving three concentric cylindrical rotating apertures is introduced which appears to be applicable to simultaneous multi-plane CT using late generation CT scanner geometries.

Line-scanned projection radiographic systems provide improved detection of low contrast objects by reducing scattered radiation and through the use of wide dynamic range detectors. However, because of the narrow source collimation required for line scanning, the maximum photon flux achievable with scanned radiography is limited. Experiments have been conducted to determine the relationship between source collimation, radiation dose, spatial resolution, and low contrast detectability using a General Electric CT/T 8800 computed tomographic scanner modified for scanned projection radiography. A contrast/detail phantom was scanned using three different source collimators and at different radiation exposures. Because the source collimation influences both the photon flux and spatial resolution, optimal collimation will depend upon the subject thickness and desired contrast and spatial resolution. Results are presented to show these tradeoffs as well as compare the performance of scanned projection radiography to conventional film radiography.

A total low dose fluoroscopy system has been developed,using a Siemens Videomed H 1023 line 25MHz television system, and a modified Matrix Videoimager, multi-image camera for recording spot films. The radiation dose during fluoroscopy has been reduced by using pulses of radiation at fluoroscopic intensities approximately 17ms. wide, storing the image on a V.A.S. video disc recorder during the pulse, and playing the stored image back through the fluoroscopic monitor. The combination of these methods has resulted in dose and cost reductions of the order of 90%, with clinically acceptable image quality. With minor modification the system could be made compatible with any commercially available high line rate, high resolution television system.

A digital video acquisition system has been built and is being expanded to evaluate the feasibility of photoelectronic radiology. The system uses specially designed photo-electronic imaging devices to eliminate the use of film as the primary recording medium for radiography. The initial studies using this system have been directed toward intravenous angiography by subtraction techniques. A system performance overview and clinical examples are presented.

The Light Beam Readout (LBRO) imaging system has been found to display a residual potential similar to that observed in the Xerographic (XR) imaging system. Both of these imaging systems utilize amorphous selenium as the photoactive element in their image receptors. It has been proposed that the phenomenon of residual potential in amorphous selenium photoreceptors is due to bulk trapping effects. An experimental method of measuring residual potential in amorphous selenium photoreceptors which may be used to test this bulk formation theory is presented.

Electroradiographic methods offer a possible alternative to conventional silver halide x-ray film/screen combinations in terms of lower-cost x-ray radiographs and more rapid development techniques. In charge transfer radiography (CTR), a semiconductor solid material exploits a photoconductive mechanism to transfer a latent electrostatic image onto a dielectric surface. The latent electrostatic image is made visible by xerographic toning. The work reported here used a sintered CdS photoconductive layer to produce large-area image receptors for CTR. Measurements of the currents generated by the incident x-ray flux show that, unlike the case for optical radiation, copper/chlorine dopants do not significantly alter the sample sensitivities. However, the sensitivity of materials thus doped is affected by crystal size and binder material. Remarkably, the sensitivity is not strongly dependent on the thickness of the photoconductor. The charge generation energy is shown to be related to the inverse of the applied electric field for both Se and CdS. An empirical expression is derived from the data that relates the exposure to the surface charge available for transference to the dielectric. Examples of paper radiographs are given, and an exposure characteristic curve for the CdS shows that the sensitivity of the sintered CdS plate is similar to normal x-ray films, but less than that of film/screen combinations.

Eight board certified diagnostic radiologists reviewed 37 tomographic examinations obtained by routine technique and dynamic technique (dynamic tomography). Based on responses to four questions, the routine tomographic technique was strongly preferred for the three different organ systems examined, (kidney, lung/mediastinum, and bone). However, selected patients benefit from the current form of dynamic tomography because of shorter examination time. Radiation exposure is also reduced. Other advantages of dynamic tomography are discussed.

The rapidly increasing usage of non-ionizing radiations--microwaves, ultrasound, ultra-violet, and lasers--for diagnostic and therapeutic applications demands an awareness of their energy levels, their potential hazards and safety thresholds. This paper provides comparative data on these characteristics, and areas of concern and further research are indicated.

Optics has had significant impact in the development and implementation of medical acoustic imaging. Principles of optical systems theory and optical instrumentation have played an important role in the conception, design and development in new ultrasonics research tools for diagnostic imaging and tissue character-ization; this is exemplified in the paper with a few acoustic imaging systems. A potential imaging system whose success relies on the interface between the optical subsystem and the acoustic subsystem is described. This system would have the advantage of operational flexibility and electronic simplicity. The present obstacles and possible solutions are discussed.

The continuing performance of radiological imaging systems has been studied by the measurement of performance indicators on two general purpose radiographic-fluoroscopic units over an extended period of time. The question of continuing performance of radiological imaging systems has become increasingly important as system complexity and cost continue to escalate and as the acquisition of replacement funding becomes more difficult. The results of performance testing and a discussion of these results are presented.

X-ray spectra and outputs are computed based on three formulations of x-ray bremsstrahlung production by Kramers, Storm, and Birch and Marshall. Published, measured spectra compare favorable with Storm and Birch and Marshall. Measured outputs (mR/mAs at 100 cm) agree best with Storm. TLD measurements of x-ray exposure agree well with predictions made using Storm over a range of techniques. Rapid calculation of output for either single phase or three phase operation is possible using coefficients from a third order fit to the data based on Storm for 40 to 120 kV and 2 to 8 mm aluminum total filtration.

Based on the author's experience, state-of-the-art cine radiographic equipment of the type used in modern cardiovascular laboratories for selective coronary arteriography must perform at well-defined levels to produce cine images with acceptable quantum mottle, contrast, and detail, as judged by consensus of across section of American cardiologists/radiologists experienced in viewing such images. Accordingly, a "standard" undertable state-of-the-art cine radiographic imaging system is postulated to answer the question of what patient exposure range is necessary to obtain cine images of acceptable quality. It is shown that such a standard system would be expected to produce a 'tabletop exposure of about 25 milliRoentgens per frame for the "standard" adult patient, plus-or-minus 33% for accept-able variation of system parameters. This means that for cine radiography at 60 frames per second (30 frames per second) the exposure rate range based on this model is 60 to 120 Roentgens per minute (30 to 60 Roentgens per minute). The author contends that studies at exposure levels below these will yield cine images of questionable diagnostic value; studies at exposure levels above these may yield cine images of excellent visual quality but having little additional diagnostic value, at the expense of added patient/personnel radiation exposure and added x-ray tube heat loading.

When a digital image of a uniform radiation source is obtained with a scintillation camera, the action of a variety of cortical receptive fields employed by the human observer can be simulated with a computer. A frequency diagram of receptor output can be generated by scanning the simulated receptor (which includes both exitory and inhibitory weightings of local image intensity). Although the measured frequency distributions exhibit a generally Gaussian shape, the dispersions are larger than predicted by Poisson statistics. The quadratic difference between measured dispersions and those predicted with Poisson statistics yields a measure of the texture contribution to noise that is interpretable in terms of small signal detection theory.

Several diagnostic X-ray imaging techniques have been developed over the last few years which are variously called digital radiography, electronic radiography, computed radiography, and digital fluoroscopy. They all have the common element of producing projection radiographic images in a digital form. Some of the advantages proposed for these systems are highly efficient use of dose, scatter reduction, ease of operation, noiseless data transmission, new types of image storage, flexible display capability to exploit the total range of detected information, and a potential for various forms of image manipulation such as edge enhancement, filtering, and subtraction.1-7 Digital radiography will succeed as a diagnostic modality only to the extent that these advantages over the well entrenched analog radiography are truly realized.

In 1972, at the first SPIE seminar covering the application of optical instrumentation in medicine, Balter and Stanton presented a paper forecasting the status of x-ray image intensifiers in the year 1980. Now, eight years later, it is 1980, and it seems a good idea to evaluate these forecasts in the light of what has actually happened. The x-ray sensitive image intensifier tube (with cesium iodide as an input phosphor) is used nearly universally. Input screen sizes range from 15 cm to 36 cm in diameter. Real time monitoring of both fluoroscopic and fluorographic examinations is generally performed via closed circuit television. Archival recording of images is carried out using cameras with film formats of approximately 100 mm for single exposure or serial fluorography and 35 mm for cine fluorography. With the detective quantum efficiency of image intensifier tubes remaining near 50% throughout the decade, the noise content of most fluorographic and fluoroscopic images is still determined by the input exposure. Consequently, patient doses today, in 1980, have not substantially changed in the last ten years. There is, however, interest in uncoupling the x-ray dose and the image brightness by providing a variable optical diaphragm between the output of the image intensifier tube and the recording devices. During the past eight years, there has been a major philosophical change in the approach to imaging systems. It is now realized that medical image quality is much more dependent on the reduction of large area contrast losses than on the limiting resolution of the imaging system. It has also been clear that much diagnostic information is carried by spatial frequencies in the neighborhood of one line pair per millimeter (referred to the patient). The design of modern image intensifiers has been directed toward improvement in the large area contrast by minimizing x-ray and optical scatter in both the image intensifier tube and its associated components.

The objective of low dose radiography is achieved by a judicious combination of proper kV selection, fast film-screen systems and beam filtration. A systematic study of filters was undertaken to evaluate the improvements that can be realized in terms of patient Entrance Skin Exposures (ESE) for chest radiographs. The Picker CD 135 Generator and the Automatic Chest Filmer with dynamic phototiming were used for the study. The kV dependence of ESE with various amounts of zinc and aluminum filtration is presented. The effect of filtration on image contrast is discussed. The variations of ESE with phantom thickness under different filtration conditions are also considered. It was found that the ESE can be reduced by as much as a factor of 1.8 ± .1 with no significant increase in tube loading.

Recent work has suggested a new approach to film/screen mammography through the use of transmolybdenum (Z > 42) fluorescent anodes. An experimental rhodium (Z = 45) anode with K emission at 20.0 - 23.2 keV has been compared to a conventional molybdenum anode. Previous spectral measurements have indicated a dose advantage for the Rh anode but no comparison of contrast has been made. A series of step wedges of different materials was imaged at several kVp and for several phantom thicknesses. The resulting data permits quantitative comparison of the radiographic contrast. At 40 kVp for an equivalent tissue thickness of 4.8 cm the radiographic contrast provided by the Rh anode is at least equal to and under some conditions actually superior to that provided by a Mo anode.

Manufacturers use either ionization chambers or photomultiplier tubes coupled to fluorescent screens as the pickup device in their automatic film exposure controls. The clinical performance of the control will be affected by its energy and exposure rate dependences. These two parameters are not usually determined independently when evaluating an automatic exposure control. By independently measuring these parameters the performance evaluation procedure can be shortened, while improving the description of performance over a wide range of tube potentials and patient thicknesses. A method to isolate the energy and the exposure rate dependences will be discussed, and measurements representing several manufacturers will be presented.

The performance of a General Electric automatic exposure control system with ionization chamber detectors was evaluated over a wide range of kVp's and acrylic phantom thicknesses for four different intensifying screens. For adult phantom thicknesses, the system kept density variations to within +0.2 optical density units for each screen. For pediatric patients, considerable excess density was observed at low mAs which was caused by stored charge in the transformer and cables. The density variations of all four screens were similiar above 70 kVp indicating that a single kVp compensation scheme is adequate for this type of detector. At 50 kVp variable compensation may be required.

A method of spatially filtering the primary x-ray beam to emphasize the middle and upper frequency structure within a patient is under investigation. In a feasibility study, a copper filter is constructed which complements the low frequency structure of a scout film. The filter is placed in the x-ray beam and a new radiograph is taken. The result is a final x-ray in which the low frequency structures are suppressed, i.e., the middle and upper frequency structures are emphasized. This provides nearly constant detection probabilities for small, low contrast structures, regardless of their location in the patient. The results obtained from several such filters are very encouraging.

A custom designed equalizing filter was used to test the potential for improving nodule detectability in chest radiographs. The filter was constructed to produce a radiograph so that the film densities over the lung field and mediastinum were approximately equal. Eighteen radiographs were prepared with small lesions placed over the lung field and mediastinal areas of a chest phantom and the films were read by four radiologists. The films prepared with the equalizing filter showed an increased true positive rate of 18% in the retrocardiac, retrodiaphragmatic and mediastinal areas and only a slight decrease of the true positive rate in the lung field (4%). The average false positive rate for films prepared with and without the filter were virtually identical.

To date, 696 patients being admitted to the Cardiac Unit have been examined by a new method-Intergated® Radiography-that adds information about cardiovascular pulsations to the conventional admission chest x-ray, without interfering with the ordinary purposes of chest radiography and without increasing radiation dose to the patient. The Intergated Radiography System (IRS) is a simple attachment for existing chest radiography units. It images the heart at end-diastole and end-systole on alternate thin strips of a single film, thus producing a composite image in which moving walls have a serrated appearance that easily permits objective measurement of wall motion while the appearance of stationary thoracic structures is substantially unaffected. Patients with coronary artery disease, aortic regurgitation and left-to-right shunts demonstrate characteristic motion patterns. IRS films compared with ventriculography and aortography show excellent correlation. Preliminary results suggest IRS may increase the diagnostic accuracy of electrocardiographic exercise stress testing for detection of coronary artery disease. Thus Intergated Radiography is a simple and objective method for screening and for following the results of therapy in patients.

The use of carbon fiber for cassette fronts and table tops has been advanced as a method of reducing x-ray exposure. We have evaluated a carbon fiber cassette by comparison to a Kodak X-omatic cassette. Although the transmission of the carbon fiber is considerably high-er than that of the X-omatic cassette, the radiation saving is limited to low kVp's. Al-though some decrease in primary/scatter ratio was measured, the overall contrast increased somewhat over most of the kVp range investigated.

Different situations are discussed in which it is of value to know radiation dose to the patient in diagnostic radiology. Radiation dose to specific organs is determined using the Handbook on Organ Doses published by the Bureau of Radiological Health of the Food and Drug Administration; the method is applied to a specific case. In this example dose to an embryo is calculated in examinations involving both fluoroscopy and radiography. In another example dose is determined to a fetus in late pregnancy using tissue air ratios. Patient inquiries about radiation dose are discussed, and some answers are suggested. The reliability of dose calculations is examined.

A modification of the Nationwide Evaluation of X-Ray Trends (NEXT) survey procedure has been used to collect data on technique and exposure values for the A/P abdomen and A/P lumbosacral spine projection for a fixed patient size. Through the use of a sample set of patient radiographs and a phantom, technique factors and entrance skin exposure (ESE) are related to a radiograph of acceptable density. An analysis of 139 surveys demonstrates the wide range of techniques and exposures associated with each projection and suggests that ranges of typical techniques and exposure values can be defined.

The Federal Food and Drug Administration (FDA) is considering development of exposure recommendations for diagnostic radiology. This paper describes one aspect of this effort: the determination of technique/exposure ranges corresponding to the range of acceptable radiographs for three commonly employed projections - P/A chest, A/P abdomen and A/P lumbosacral spine. The Radiology Departments of nine Manhattan medical facilities will each provide 100 cases for each projection for patients in the thickness range of 21-25cm. Data collected for each of these 2700 radiographs will include: patient x-ray technique (kVp, mAs), patient thickness, image receptor data (film, screen, grid), patient entrance exposure, beam HVL, focal spot size, image system speed and contrast, and the radiologist's assessment of radiographic quality. The design of the study and preliminary data will be presented. The study will be completed in 1981.

It has been proposed that single skin entrance exposure values be recommended for specific radiographic examinations. Numerous factors, however, contribute to large excursions in the measured skin entrance exposure (ESE) received by patients even in idealized conditions. Excluding the obvious effects of generator miscalibration, suboptimal film processing, film/screen spectral mismatches, et cetera, large variations in ESE are observed, even though clinically satisfactory radiographs are obtained. In particular, variations in ESE can be attributed to: selection of the film/screen combination, use of grids, patient thickness and tissue composition, kVp settings and the filtration employed. Data were taken at our institution for several representative examinations under various conditions. Lithium Fluoride TLD ribbons were placed on a series of patients undergoing PA and Lateral chest radiographs and PA abdomen radiographs with barium for Upper GI series. The patient ESE data observed at our institution, together with data taken with an anthropomorphic phantom will be presented. These data will be used to demonstrate that although a single ESE guideline may be appropriate under a specific set of conditions, a single value may be either too restrictive for a slower system or excessive for a faster system.

Due to an avid interest to reduce unnecessary patient exposure, the Illinois Department of Public Health developed exposure limits for certain medical and dental radiographic examinations. These limits were determined by evaluating and analyzing actual patient exposure measurements conducted at medical and dental radiographic facilities throughout the state. A tremendous range of exposures existed for the same examination, and it became evident that patients were being needlessly overexposed to radiation. Marked decreases in exposure have resulted since the adoption of these limits as a regulatory method to reduce unnecessary patient exposure.

Slit radiography is a highly efficient technique for rejecting scattered radiation in x-ray imaging. In slit radiography, a large-field image is recorded by scanning a narrow x-ray beam across the object, so that all parts of the image are formed by a small x-ray field. Several investigators, including ourselves, have constructed slit radiography machines to test this principle. In general, slit radiography has been shown to be highly effective for eliminating scattered radiation, capable of reducing scatter intensity to less than 20% of the recorded x-ray beam. Problems with the technique also have been revealed, including: 1) the need for a very precise and mechanically rugged collimating apparatus to shape and scan the slit x-ray beams, 2) exceptionally high tube loading and/or exposure time requirements due to inefficient utilization of x-rays in forming narrow slit beams, and 3) insufficient latitude of currently available x-ray film types for recording the high contrast images generated by the slit technique, especially in procedures that already are latitude limited, such as chest radiography. The potential for practical implementation of slit radiography techniques, and their impact on diagnostic accuracy in light of these problems is discussed.

A new radiopharmaceutical, 68Ga-iron hydroxide colloid, for hepatic imaging by positron emission tomography (PET) was prepared from the eluate of a "Ge-68Ga solvent extraction generator. In rats, 84% of the administered dose of colloid localized in the liver and 4.6% accumulated in the spleen. Initial imaging studies in normal dogs showed close correspondence of the findings by PET and transmission computed tomography (CT). PET with 68Ga-colloid was performed in 10 patients with hepatic metastases demonstrated by conventional scintigraphy with 99mTc-sulfur colloid. All focal defects noted on the conventional scintigrams were easily identified and generally seen more clearly by PET. In one patient, lesions not identified on the initial 99mTc-sulfur colloid images were demonstrated by PET. The positron tomographic images were compared with those obtained by CT in 7 patients; the two studies showed comparable findings in 5 patients, whereas PET more clearly showed multiple lesions in 2. Our results suggest that PET is a suitable technique for obtaining high-contrast, cross-sectional images of large abdominal organs. Emission computed tomography with positron-emitting radionuclides shows promise as an important new tool for clinical research (1-4). Unfortunately, wide clinical application of positron-emission tomography (PET) is presently limited by the need for an expensive, hospital-based cyclotron facility and highly trained professional and technical personnel to synthesize the radiopharmaceuticals labeled with the very short-lived radionuclides 11c, 13N, 150 and 18 F that are employed most commonly in such studies. These difficulties may be circumvented in part by the use of a simple generator system that produces the positron-emitting radionuclide 68Ga (T1/2 = 68 min) from the long-lived parent 68Ge (T1/2 = 275 days) (5-7). A large number of radiopharmaceuticals of potential clinical interest may be prepared readily from the eluate of such a generator (6,8-11). In the present study, we have prepared and characterized a new radiopharmaceutical for liver imaging, 68Ga iron hydroxide colloid, and have evaluated this agent in patients with hepatic tumors by imaging with a positron tomographic scanner.

Heavy particle radiography is a newly developed noninvasive low dose imaging procedure with increased resolution of minute density differences in soft tissues of the body. The method utilizes accelerated high energy ions, primarily carbon and neon, at the BEVALAC accelerator at the Lawrence Berkeley Laboratory. The research program applied to medicine utilizes heavy-ion radiography for low dose mammography, for treatment planning for cancer patients, and for imaging and accurate densitometry of skeletal structures and brain and spinal neoplasms. The presentation will be illustrated with clinical cases under study. Discussion will include the potential of heavy-ion imaging, and particularly reconstruction tomography, as an adjunct to existing diagnostic imaging procedures in medicine, both for the applications to the diagnosis, management and treatment of clinical cancer in man, but also for the early detection of small soft tissue tumors at low radiation dose.

A significant portion of the dose received by a patient during a radiographic x-ray examination results from absorption of the beam in the various patient support structures through which the beam must pass. New materials are becoming available that offer reduced absorption of x-rays while providing the strength and rigidity required. One such material is CFM. Comparative measurements on CFM and conventional materials (CM) under laboratory and practical conditions are presented. These measurements address the effect of these materials on patient dose, scattered radiation, and contrast.

An evaluation of graphite and conventional radiographic table tops and screen-film system cassettes under clinical conditions is presented in terms of potential radiation dose reduction, cost, and image quality.

The application of digital subtraction angiography (DSA) for imaging of the human cardiovascular system is described using new, specifically designed research equipment. The system provides the ability to obtain real time or off-line subtraction images from digital data. Results show that DSA, performed by the intravenous injection of contrast material, can provide images comparable to conventional angiography in multiple areas of the cardiovascular system and on an outpatient basis.

The use of detailed performance specifications as an integral part of the purchase agreement for radiologic equipment is gaining increasing acceptance in the radiologic community. This approach to communication between the purchaser and the vendor of radiologic equipment has the potential to provide improved correlation between the capabilities of the equipment and the clinical needs to which it is applied. The approach also has the potential to create additional work for both the purchaser and the vendor with little benefit to either party, if the performance specifications and related documents are not developed carefully and methodically. As a guide to this development, the tables in this document are provided. The procedures in these tables should be followed sequentially, and the tables themselves are listed in the order in which they should be encountered in the purchase process.

A new medical thermographic instrumentation system provides for direct measurement of human body surface temperatures and automatic on-line diagnostic interpretation by a microcomputer. This system directly measures 128 surface temperature areas with an infrared scanner array of 64 thermopile sensors. This thermal data is fed to a microcomputer which executes a diagnostic pattern recognition algorithm and prints out a report which summarizes the measurement data. Initial results indicate both a sensitivity and specificity of 80%. However, 80% of the women with cancer had Stage II or greater disease. An improved design of the initial prototype has been constructed.

The benefits and structure of the inhouse service program at the University of Alabama in Birmingham are reviewed. The program, when analyzed and compared to contracted service, realizes a savings of from $190,000 to $250,000 per year. Scaling the program to smaller hospitals indicates that inhouse service becomes cost effective when 11/2 million dollars or more of equipment is involved. For remotely located hospitals the cost effective equipment figure may be somewhat less.