HomeRadiologyVol. 91, No. 6 PreviousNext Book ReviewsRoentgen Diagnosis in Five VolumesJohn A. CampbellJohn A. CampbellJohn A. CampbellPublished Online:Dec 1 1968https://doi.org/10.1148/91.6.1240MoreSectionsPDF ToolsAdd to favoritesCiteTrack CitationsPermissionsReprints ShareShare onFacebookXLinked In Article HistoryPublished in print: Dec 1968 FiguresReferencesRelatedDetailsRecommended Articles RSNA Education Exhibits RSNA Case Collection Vol. 91, No. 6 Metrics Altmetric Score PDF download

HomeRadiologyVol. 90, No. 5 PreviousNext Announcements and Book ReviewsBook ReviewsError and Variation in Diagnostic RadiologyJohn A. CampbellJohn A. CampbellJohn A. CampbellPublished Online:May 1 1968https://doi.org/10.1148/90.5.1033bMoreSectionsPDF ToolsAdd to favoritesCiteTrack CitationsPermissionsReprints ShareShare onFacebookXLinked In Article HistoryPublished in print: May 1968 FiguresReferencesRelatedDetailsRecommended Articles RSNA Education Exhibits RSNA Case Collection Vol. 90, No. 5 Metrics Altmetric Score PDF download

In our continuing study of median craniofacial anomalies, we have found it convenient to subdivide cases on the basis of the interorbital distance. Orbital hypotelorism in combination with a proboscis or with a flat nose and median cleft lip reliably predict a brain that has failed to divide into cerebral hemispheres (holoprosencephaly) (3). The close correlation between face and brain in these patients prompted us to study the median facial anomalies associated with orbital hypertelorism (1). This communication reiterates the results of that study with emphasis on the roentgen characteristics. Classification and Roentgen Description Orbital hypertelorism occurs repeatedly with six other median facial anomalies: (a) low “V”-shaped frontal hairline, (b) cranium bifidum occultum frontalis, (c) primary telecanthus (lateral displacement of the medial canthi relative to the pupils), (d) median cleft nose, (e) median cleft prolabium2 and premaxilla,3 (f) median cleft secondary palate. In addition to these abnormalities, some patients have frontal lipomas, dermoids, or teratomas. Of this group of anomalies, orbital hypertelorism, cranium bifidum occultum frontalis, median cleft nose, and median cleft prolabium and premaxilla were considered basic in further subdividing cases into four facial types. This classification is based on a study of 25 cases, including 8 personal ones (1). The findings in Type 1 fades (3 of 25 cases) are: (a) orbital hypertelorism, (b) complete cleft of the nose, (c) absence or marked hypoplasia of a cleft prolabium and premaxilla, and (d) cranium bifidum occultum frontalis. In addition, in this facial type the palatine processes of the maxillary bones (secondary palate) may be widely separated. Case I: U. M., a 1-month-old Negro male, had the most severe form of the median cleft face syndrome encountered and probably the most severe form compatible with life (Fig. 1). Evaluation at thirteen months of age suggested mild mental retardation. Skull roentgenograms disclosed a wide cleft (50 mm) between the bony orbits, maxillary bones, and ethmoid elements, with the cleft extending posteriorly to the sphenoid bone. The primary direction of the orbital axes was lateral. The premaxilla and palatine processes of the maxillary bones could not be identified. The anterior clinoid processes and the sella turcica were normal in appearance. Ossification of the frontal bones was deficient above the cleft. The head size was normal. Case II: C. H. was a 1-day-old male infant. Frontal and lateral skull roentgenograms showed marked orbital hypertelorism (31 mm) (Fig. 2). The maxillary bones were separated with the interspace bridged by “ethmoid filler.” The cleft in the secondary palate extended superiorly into the ethmoid area. The premaxilla could not be identified. There was an extensive defect in ossification of the frontal bones with a median bar of bone. The anterior clinoid processes and the sella turcica appeared normal.

An excellent recent publication has catalogued more than 100 syndromes affecting chiefly the head and neck (7). These sign and symptom complexes occur infrequently, but as a group they represent an important segment of medicine. Their recognition and correct diagnosis are important, particularly for the prognostic and possible genetic implications. In addition, clear delineation of a disease entity and its clinical spectrum is a requisite for serious consideration of etiology. Because of the similarity in facial appearance of some patients with different syndromes, confusion has developed which has tended to retard progress in correctly categorizing several disease entities. This communication is designed to emphasize the role of roentgen analysis as a complement to the clinical diagnosis of these rare syndromes, features of which, particularly in infancy, may appear to be limited largely to the head. Two cases, one of the oculodentodigital syndrome, the other of the Hallermann-Streiff syndrome, will be reported with emphasis on the roentgen features in infancy. In spite of the clinical similarities of these diseases, a clear roentgen differentiation can be made. Case Reports Oculodentodigital Syndrome K. B., a 5-day-old white female, was admitted to the Indiana University Medical Center because of low birth weight (4 lb.) and peculiar facies. She was the second of nonidentical twins; the other, a female, weighed 5 lb., 3 oz., and was apparently well. The mother, aged thirty-one, had no prenatal care, but there were no apparent complications during the period of gestation. Menstrual history was vague, so the exact gestational age could not be determined. The mother's health was good, and there was no apparent evidence of physical abnormality. The only drugs reportedly used during gestation were proprietary analgesics, cold remedies, and the like, although contraceptive jelly had been employed intermittently at about the time of conception. Legal paternity of both twins was excluded by genotyping; there are reported to be at least 14 half-siblings, 5 maternal and 9 paternal, living and well. There is no known consanguinity between the true parents. Admission vital signs were: pulse 132, respirations 32, temperature 98.2°F., weight 3 lb., 12 1/2 oz. The cry was feeble, and the patient responded poorly to external stimuli. The skin had a yellowish tint. The occipitofrontal circumference was 29.8 cm. Physiognomy was unusual with a “bird-like” nose, small lower jaw, but no obvious ear abnormalities (Fig. 1, A and B). The palpebral fissures were small and difficult to open. There was bilateral microphthalmia, the corneas each measured 6 mm in diameter, and the sclerae were white. The skin was extremely thin, especially over the skull where the veins were easily seen. The chest and abdomen were not remarkable. Cutaneous webbing of the fourth and fifth fingers of both hands was present, partial on the right.

Just as phototiming enhances the usefulness of spot-filming by eliminating exposure error, some means of assuring proper film density is a prerequisite for the broader clinical application of cineradiography. In the past, manual control of milliamperage by a millivolt monitor of the image tube current (Philips Co.) and a pre-read microammeter measuring the phototube current of a brightness-stabilizing circuit (Westinghouse) were used to establish correct cine film exposures in this work. Later, improved brightness stabilizers were employed in an attempt to maintain constant light levels for recording anatomical areas of varying opacity. While these devices, if properly used, permit one to achieve satisfactory film exposures, they are tedious to adjust, and the task of obtaining reproducible results under all conditions of cine operation requires a rather complicated procedure on the part of the radiologist. There is need for a simple-to-operate device which will assure constant film density at all film speeds, framing rates, and camera adjustments with either manual or automatic control of kilovoltage and milliamperage. Obviously, the ideal solution would be to place a photocell in the camera magazine which could monitor the light intensity passing through the shutter. The predetermined potentiometer setting of this phototube would limit the exposure time to that consistent with proper film exposure just as in spot-film phototiming. This placement of the photocell has not yet been successfully accomplished because of the physical problems incident to sampling the light intensity in the focal plane of the camera lens. Consequently, at the present time it is advisable to monitor samples of varying light intensity emerging from the center of the output phosphor and attempt to correlate these with the exposure requirements of the film in the camera. This method will be imperfect unless it compensates for the variation in the open time of the camera shutter at different filming speeds which permit the afterglow of the image-tube phosphor to contribute to the film exposure. Another shortcoming of systems which scan the full field of the output phosphor is their failure to adjust the exposure factors for changes in the field size during filming, resulting in over- or underexposure of the film whenever the field size is significantly changed by the radiologist. This paper describes an exposure meter circuitry which embodies simplicity and reliability of operation and is designed to operate on a semiautomatic basis. It offers a predetermination of proper exposure density and at the same time permits the radiologist to control manually the variable factors influencing cine exposures.

Cine cameras, both 16- and 35-mm. size, record their images in a rectangular frame, the height measuring the shortest diameter (Fig. 1, A). As cine-radiographic images are circular in shape, it is apparent that the maximum diameter which may be recorded is restricted to the height of the rectangular frame. In order to obtain a reasonable compromise, it is frequently necessary to overframe the image (Fig. 1, B). This may result in a loss of as much as 50 per cent of the vertical dimension of the image, whether on 16-or 35-mm. film. After subtracting the space utilized by side perforations, the useful frame size of 16-mm. film is 7 1/2 × 11 mm., and 18 × 26 mm. for 35-mm. film. Some gain in the width of the frame may be achieved through the use of single perforated 16-mm. film by utilizing the unperforated edge. However, while this adds approximately 2 mm. to the width of the frame it obviously is of no great value in cineradiography as it does not extend the critical vertical diameter of the conventional frame. The only method of recording a circular image with a vertical dimension equal to the maximum horizontal dimension is to provide a square frame instead of the conventional rectangular one. Square-frame photography obviously requires certain modifications of the currently available films, cameras, and projectors. This paper describes a method of achieving this on 16-mm. recordings. Film Modification In order to expose a square 16-mm. film frame, it is necessary to move the film an additional half frame into the camera or projection aperture with each frame advance. This requires a film which offers additional perforations spaced halfway between the usual perforations of regular 16-mm. film (Fig. 1, C). This type of perforated film is readily available in the 16-mm. size which is manufactured for use in 8-mm. cameras and is ideally suited for square framing.3 Single perforated film is preferable as it allows extended coverage in the horizontal as well as the vertical frame dimensions by utilizing the unperforated edge (Fig. 1, D). Camera Modifications The following modifications are necessary in the camera. Camera lens focal length varies proportionately to the size of the area filmed according to the formula: Consequently, it is necessary to lengthen the usual 25-mm. focal length used on the standard 16-mm. camera to about 40 mm. on a square-frame unit. The aperture on the standard camera magazine must also be lengthened and slightly widened to allow the full square image to be projected on the film (Fig. 2). The film advance mechanism must also be modified so as to center the square area in the film gate. This may be done by increasing the shuttle gear ratio so that a conventional frame and a half is moved forward with each exposure. In addition, the take-up reel is also provided with a greater pulling ratio to allow for collection of the greater length of film advanced.

The accurate determination of specific chamber size is an integral part of the radiographic examination of the heart and great vessels. The left atrium and right ventricle can be evaluated with a high degree of accuracy by the utilization of well known and valid criteria. The posterior border of the left atrium is in juxtaposition to the esophagus, and enlargement of that chamber displaces the barium-filled esophagus posteriorly and to the right. This finding is best demonstrated in the lateral and right anterior oblique views. Enlargement of the right ventricle rotates the heart in a clockwise direction. On the postero-anterior view, the apex is usually elevated. Because the right ventricle is an anterior chamber, enlargement is well demonstrated in the right anterior oblique and lateral views as increased convexity of the anterior cardiac border and a decrease in the retrosternal space. Classically, the left ventricle enlarges posteriorly, downward, and to the left, as shown in the postero-anterior, lateral, and left anterior oblique views. A number of cases, however, with hypertrophy of the left ventricular wall without dilatation of the left ventricular cavity do not present classical radiographic findings, and in certain cases it is difficult to differentiate left ventricular from right ventricular enlargement. The electrocardiogram adds useful information for solution of this problem. The right atrium is the most troublesome chamber to evaluate. It overlaps the right ventricle and when that chamber is also enlarged (as is true in most conditions causing right at rial enlargement) , the determination of right atrial size is attended by considerable difficulty. The purpose of this paper is to re-evaluate the accuracy of various radiographic crite ria for determining the size of the right atrium (1–7) and to present a simple fluoroscopic method which has proved, in the authors' hands, to be extremely accurate. Method of Study Two hundred patients were chosen for this study. One hundred had uncomplicated ventricular septal defects and 100 had uncomplicated atrial septal defects. All cases were proved by cardiac catheterization and cinecardioangiography. Radiographs in the usual four projections were available for all. Cinecardioangiograms in at least the left anterior oblique view were technically adequate for study in 86 patients with ventricular septal defects and 89 patients with atrial septal defects. The group with ventricular septal defects were considered to have right ventricular enlargement only, while those with atrial septal defects were considered to have both right ventricular and right atrial enlargement. An additional 100 patients with no signs or symptoms of cardiac disease were examined fluoroscopically in the left anterior oblique projection for further evaluation of pulsation of the right cardiac border.

In the practical application of cineradiography, the radiologist needs a reliable and speedy determination of proper film exposure. Failure to achieve this greatly hampers the full realization of the clinical potential of this method of studying dynamic motion and restricts the extension of its usefulness to other areas of diagnostic roentgenology. As with spot-film radiography, some type of phototiming would be a desirable means of assuring correct film density for any combination of emulsion speed, framing rate, object density, and radiation exposure factors. Unfortunately, phototiming as we know it in conventional radiography cannot be readily achieved without a radical change in commercially available cine cameras. The compact design of these cameras does not provide sufficient space for the insertion of even a small conventional photocell into the area of the filming aperture. It is conceivable that a tiny selenium (cadmium sulfite) type photosensitive crystal could be placed in the camera, but this type of pick-up is not linear in its response to the flickering light passing through the camera shutter, and reproducible results would be difficult to obtain. Furthermore, several variables not encountered in conventional radiography tend to complicate the phototiming of cine exposures. First, the actual exposure time in cineradiography is governed by an x-ray pulse synchronized with the opening time of the camera shutter or by the angle aperture of the camera shutter alone. Since the exposure time is therefore always of the same duration for a given camera speed, the intensity of exposure is varied by changing kilovoltage or milliamperage rather than the time factor. Also, since the image is formed by the intensifier tube, the response of this instrument in terms of brightness gain and contrast resolution to a given input change in the quality and quantity of radiation must be anticipated by the controlling device. In addition, the fact that films of different speed and grain are desirable for a variety of recording requirements makes it necessary to have an easily adjustable exposure control properly corresponding to the film response. Along with these requirements is the need for a brightness stabilization circuit to maintain constant intensity of the image during fluoroscopy and cineradiography of anatomical parts of varying thickness and density. Since brightness stabilization is an optional extra provided by equipment manufacturers, some machines are furnished without this circuit. The majority of stabilizers operate from a phototube sampling the light intensity from the output phosphor of the image tube, while others regulate brightness by keeping the milliampere current proportional to the electron flux across the image tube.

In the course of the clinical interpretation of well over 600 angiocardiograms at Indiana University Medical Center during the past twelve years, the aortic configuration was found to be the key to the recognition and differentiation of many primary types of congenital and acquired cardiac lesions. With the use of selective cinecardioangiography during the past three years, the importance of aortic size in differential diagnosis was even more impressive. This knowledge adds specificity to the plain film findings and narrows the field of diagnostic possibility by exclusion of inconsistent aortic findings. To utilize the aortic configuration to the greatest diagnostic advantage, it is necessary not only to recognize the various morphologic pictures which occur in different lesions, but also to appreciate the frequency and reliability of the roentgenologic evidence as it pertains to the type of altered hemodynamics involved. It must be remembered that the normal aortic configuration of children and adults is not the same. As a patient ages there are continuous degenerative changes in the aortic wall so that there is a relative increase in aortic size. A normal aortic configuration in a patient fifty years of age would be definitely abnormal in a child of ten. If the normal gradual increase is taken into consideration, the aortic size can be closely related to the volume and pressure of the blood entering the systemic circuit. In evaluation of the usefulness of the size, shape, and specific aortic contours in the diagnosis of congenital heart disease, a systematic review of the plain films of 610 cases of 12 commonly encountered lesions was carried out, and the results were tabulated. The findings were correlated with the anatomical observations made on angiocardiography, at surgery, or at autopsy. This paper reports our experience in this endeavor. Left-to-Right Shunts Patent Ductus Arteriosus: The ductus arteriosus is a communication between the inner surface of the aorta and the left pulmonary artery. If this vessel does not close at birth, there is a shunt of blood from the higher-pressure aorta to the lower-pressure pulmonary artery. The amount of flow depends on the relative pressures of the systemic and pulmonary systems and the anatomical size of the ductus, causing enlargement in the caliber of the ascending and transverse aorta. There is an increased blood flow through the aorta proximal to the mouth of the ductus. There may be, also, a localized conical tenting of the aorta at the site of the ductus. This was first described by Rokitansky (1) in 1852 but was of little practical significance until the advent of angiography, when it was rediscovered by Jönsson and Saltzman (2). It is known as the infundibulum of the ductus and can be visualized in 25 to 50 per cent of cases of patent ductus (3,4) (Fig. 1). It is seen on the posteroanterior radiograph as a convex bulging of the lateral margin of the aorta immediately below the knob.

In obtaining over 300 cinecardioangiograms on 16-mm. film in the past two years, we have had an opportunity to evaluate the various technical details which may critically influence the diagnostic yield of this procedure. Most of these factors have been appreciated as the result of discovering technical errors, correction of which resulted in an important improvement in the diagnostic quality of the examination. Some materials and devices have been tested experimentally in the development of a more successful procedure. We are utilizing both Westinghouse and Philips 5-inch image intensifiers equipped with 16-mm. motor-driven cinecameras for this work. These units have been adapted to standard fluoroscopic tables and 300-ma, 120-kv generators. Focal Spot Size: The unsharpness factor of the film image can be slightly reduced by using a 0.3-mm. instead of a 1.0- or 2.0-mm. focal spot. This is due to the absolute gain in definition of the image itself. The improvement is particularly noticeable because the tube-table top distance is short (18 inches), and the object-screen distance is often unavoidably long (up to 10 inches), so that any unsharpness is appreciably magnified. The extent to which the 0.3-mm. focal spot may be used for cine work is limited by the local heat capacity of the face of the anode. For focal spots of 1.0 mm. or more, the tube rating is limited by the total heat capacity of the anode. There is no reason, however, to utilize a focal spot larger than 1.0 mm. The published tube rating of the 0.3-mm. focal spot confines its use to children. We frequently exceeded this limit by several fold, and the life of our first two tubes was limited to approximately 150 cases each, plus several hundred hours of fluoroscopy. This tube life at first seems exceedingly short, but by calculation each tube actually made over 150,000 individual exposures. A 0.5 and 1.0-mm. focal spot combinaation on a high-voltage tube (Dynamax 46) appears most useful at present for cine work. This has recently been made available by the Machlett Company (1). Image Magnification Factors: Since the field size is a critical limitation in cineradiography, the image magnification factor is important. Modern fluoroscopic tables utilize an 18-inch target-table top distance. The distance from the target to the midplane of a 20-cm. patient would be 56 cm. The distance from the midplane to the front of the image tube would be 10 cm. Actually, on a currently available 5-inch tube (Philips), the input phosphor is 5.9 cm. above the front of the image tube (2); thus the midplane-input phosphor distance would be 15.9 cm. The per cent of magnification can be calculated and is recorded in Table I for various target-table top distances. For an 18-inch target-table top distance the magnification would be 28 per cent. Thus, only 3.6 inches of the midplane structures would be recorded on a 5-inch image tube.

The recent literature concerning sulfapyridine and sulfathiazole renal calculi, causing anuria with at least six deaths reported in a summary of the literature out of 29 cases by Kawaichi and Rogers,1and 2 deaths reported by Lindner and Atcheson,2testifies to the lethal qualities of these drugs. That the new drug sulfadiazine should be devoid of these serious complications, as was suggested by experimental and clinical reports, was optimistic to say the least. Trevett, Nelson and Long3found only 4 instances of hematuria, with renal colic in only 2 of 125 patients given sulfadiazine. Billings and Wood4treated 105 patients with pneumonia using sulfadiazine intravenously and orally without a single renal complication, i. e. hematuria, renal colic, oliguria or anuria. Borst5states that the crystals of the acetylated and active free forms of sulfadiazine are extremely soluble, so that renal damage is nil. Finland, Strauss