American Journal of Roentgenology

This review will discuss imaging of the chest in patients with pulmonary emphysema. Imaging findings must be related to the structure of the lung because emphysema is defined in anatomic terms. Accordingly, we first review the anatomic definitions of emphysema and its consequences and then review the imaging findings, with emphasis on CT, in patients with this disease. The more severe the morphologic emphysema, the more likely a radiographic diagnosis will be made, no matter what criteria are used. The criterion of arterial deficiency is specific but insensitive. The criteria used to assess overinflation are sensitive but not specific. CT can be used for both qualitative and quantitative assessment of emphysema. The presence and extent of emphysema can be determined by visual assessment of areas of abnormally low attenuation or by objective quantification based on the attenuation values. Statistically significant correlations between emphysema and CT findings have been shown in numerous studies, but mild morphologic emphysema may be missed by CT, and occasionally CT scans give false-positive findings. In patients with moderate to severe emphysema, the severity of emphysema is underestimated on the basis of CT findings by a factor of approximately three when compared directly with results of pathologic examination of lung specimens. In spite of these limitations, CT is the best way of recognizing emphysema in living patients and probably has a significant role in recognizing localized emphysema that is amenable to surgical treatment.

Why should radiologists be interested in technology assessment and outcomes research? The primary answer relates to optimizing the use of imaging examinations in daily practice such that this results in the most effective, efficient, and highest quality patient care possible. In formal outcomes research, this is measured in terms of quality of life and perceptions of patients' well-being. Thus, the primary answer is that radiologists need to become better informed and knowledgeable about research methods and be able to critically review the results in the literature of technology assessment and outcomes research regarding the use of imaging examinations. This will enable radiologists to become more effective consultants for their referring physician colleagues by optimizing the use of imaging examinations to affect patient outcomes. Radiologists also will be more effective participants in guidelines task force groups.

Nosocomial (i.e., originating or taking place in a hospital) pneumonia is the leading cause of death from acquired nosocomial infections. The estimated prevalence of nosocomial pneumonia in intensive care units ranges from 10% to 65%, with fatality rates of 13-55%. Ventilator-associated pneumonia (VAP) specifically refers to nosocomial pneumonia in a mechanically ventilated patient that was neither present nor already developing at the time of intubation (i.e., clinical evidence of VAP occurring > 48 hr after intubation). During the past decade, some studies have suggested that VAP can be an important determinant of out-come for critically ill patients requiring mechanical ventilation. Recent investigations have provided new insights into the pathogenesis of VAP, and improved techniques have been developed for its diagnosis. Most important, emerging clinical data now suggest that new management strategies for VAP, including more specific indications for antimicrobial use, may significantly improve patients' outcomes.

Current and potential roles for MR imaging in the management of breast disorders are reviewed along with the specific technical requirements for each application. Major topics include (1) evaluating breasts before biopsy to reduce the number of surgical biopsies for benign lesions, (2) staging of breast carcinoma in breast conservation candidates, (3) evaluating breasts with inconclusive conventional imaging, (4) coordination of minimally invasive surgery, and (5) evaluating silicone implant integrity.

Chest radiography is the most frequently performed radiologic imaging study and also one of the most challenging. The technical aspects of the conventional chest radiographic examination have been studied extensively, and the effects of various parameters on the quality of chest radiographs have been described. However, new approaches to image acquisition and display have been introduced in the last decade, and many circumvent the limitations of conventional screen-film studies. Thus, a number of practical issues must be considered in the selection of chest radiographic equipment for modern clinical practice. In this review, we describe several recent advances in both conventional and digital technologies that can improve the diagnostic quality of chest radiographs.

Established noninvasive methods for assessing myocardial ischemia have limitations that might be overcome by MR imaging. We investigated MR myocardial perfusion imaging and MR ventriculography, before and after dipyridamole-induced stress, to determine whether the superior spatial and temporal resolution of MR imaging has advantages for the evaluation of myocardial ischemia.

Eighteen patients with symptoms suggestive of myocardial ischemia were examined by use of MR perfusion imaging and MR cineangiography before and during dipyridamole-induced stress. Multiplanar gradient-echo MR cineangiography and cardiac gated fast low-angle shot (turbo-FLASH) MR imaging during injection of gadopentetate dimeglumine were used. Results were compared with findings from perfusion scintigraphy and coronary arteriography.

The accuracy of the combination MR technique for detecting myocardial ischemia was similar to that of scintigraphy. No significant difference was found between the MR technique and scintigraphy for detecting segments of myocardium supplied by stenosed coronary arteries (> 70% reduction in diameter, as determined by coronary arteriography). The sensitivity of the combination MR technique for angiographically detecting significant coronary artery narrowing was 92%, and the specificity was 100%. For scintigraphy, the sensitivity was also 92% and the specificity was 100%.

Initial results indicate that a combination of stress MR myocardial perfusion imaging and MR ventriculography is feasible and that this technique can detect myocardial ischemia with an accuracy similar to that of scintigraphy. This technique may make more complete noninvasive assessment of myocardial ischemia possible.

This pictorial essay illustrates the pre- and postoperative findings seen in patients with aortic dissections on contrast-enhanced turbo-fast low-angle shot (FLASH) MR images obtained in the plane of the aortic arch. Contrast-enhanced images provide morphologic and functional information not normally available with conventional spin-echo (SE) MR imaging or dynamic CT. Preoperative examination of the acute dissection is often troubled by cardiovascular insufficiency and motion artifacts. Therefore most of our patients were examined postoperatively. The main reason for postsurgical imaging is the evaluation of the flow in the different lumina and the detection of complications (i.e., aneurysms or progress of dissection).

Chest and abdominal CT scans using 1.0-sec scan times are often limited by motion in patients who are unable to hold their breath. With our scanner we can obtain images in 0.6 sec (partial scan) that use data from 225 degrees instead of the 360 degrees used for 1.0-sec scans. The purpose of this study was to assess whether the quality of images of the chest and abdomen in patients on mechanical pulmonary ventilators who could not breath-hold could be improved if images were taken using a scan time of 0.6 sec rather than 1.0 sec.

Thirty patients who were being treated with mechanical pulmonary ventilation with indications for chest or abdominal CT or both were scanned with a scan time of 1.0 sec. At the end of the examination, additional 0.6-sec images were taken at three or four levels. Images obtained with these two scan times were filmed at similar levels and windows, and anatomically matched levels were compared. The images were masked and independently evaluated by three radiologists for motion, noise, artifact, and overall image quality. Each parameter was rated using a scale of 1 to 4. Using Kendall's tau correlation, there was no significant difference between the radiologists in the grading of individual parameters or of overall image quality. Therefore, the average of the scores of the three radiologists was used. Statistical analysis was done using repeated measures multivariate analysis of variance.

Images obtained in 0.6 sec had significantly less motion (p < .001) but more noise (p < .001) than those obtained in 1.0 sec. We found no statistically significant difference in artifacts between the two scan times. Overall image quality was judged to be significantly better on scans obtained in 0.6 sec than on scans obtained at 1.0 sec (p < .001), in spite of the greater noise on 0.6-sec images. The higher quality of 0.6-sec images was most noticeable for lung window settings.

Our results indicate that scans taken in 0.6 sec (partial scans) provide better quality images than those obtained in 1.0 sec in patients being treated with mechanical pulmonary ventilation who cannot breath-hold. This technique may be useful not only in this population but in all patients who are unable to cooperate with breath-holding instructions.

A hyperlucent thorax on plain chest radiography indicates a decrease in the radiographic density of the thorax, which can be caused by intra- or extrapulmonary diseases. The purpose of this study was to assess the prevalence and mechanisms that may be responsible for unilateral hyperlucency of the thorax after neck dissection and to determine if atrophy of the trapezius due to the transection of the accessory nerve is a cause of hyperlucent thorax.

Differences in the radiographic density between the right and left lung were evaluated and correlated with transection of the accessory nerve in 21 patients who had had a radical or a modified neck dissection for a malignant tumor of the head and neck. Twenty-eight neck dissections were performed on these 21 patients (seven had a simultaneous bilateral neck dissection). In 14 of the 21 patients, the accessory nerve had been severed during the neck surgery. In six patients, mechanisms responsible for a hyperlucent thorax were investigated with follow-up thoracic CT scans.

In the radiographs, eight patients had a hyperlucent thorax on the side of the neck dissection. In all of these cases, the accessory nerve on the side of this neck dissection had been severed during a radical or a modified neck dissection. Prior to surgery, no such hyperlucency was noted. CT scans showed atrophy of the denervated trapezius muscle.

Our findings show that atrophy of the denervated trapezius muscle after neck dissection is a cause of unilateral hyperlucent thorax on plain chest radiographs. Therefore, this finding should be anticipated as a postoperative change in patients who have had this surgery.

We undertook a study to evaluate the usefulness and the cost of our policy of obtaining all previous mammograms, whether done at our facility or another, for purposes of comparison with current mammographic examinations.

We reviewed 1432 randomly selected screening mammography examinations from the year 1992. Information collected included whether there were previous mammograms, whether they could be obtained, the clinical impact of the comparisons, and the time and cost of obtaining the earlier mammograms.

Of the 1432 cases, 1245 women (87%) had had previous mammograms: 971 (78%) of these were done at our institution (all but 12 were available for comparison) or were brought in by the woman at the time of her scheduled examination; 274 (22%) were done elsewhere and were not available at the time of the examination. Of the latter 274 examinations, 140 (51%) could not be obtained despite vigorous efforts. Clinical management was affected as a result of comparison in 35 of the 1093 cases in which previous mammograms were obtained: seven biopsies were performed (four benign, two malignant, one pending), 10 imaging workups or follow-ups were initiated, two biopsies were avoided, and 16 imaging workups or short-term follow-ups were avoided. The two cancers that were detected as a result of obtaining previous films included one manifested by an increase in calcifications and one by a neodensity that was recognized in retrospect in an otherwise normal mammogram. The average labor and postage cost for each outside mammogram requested and received was $12.52; if the film was requested but not received, the average labor cost was $5.33.

When interpreting mammograms, a comparison with previous examinations has a positive impact on clinical management and cancer detection in a limited number of cases. The overall costs and time involved in obtaining previous mammograms from other facilities are substantial. The previous mammograms cannot be obtained in all cases, and in our study only 50% of the mammograms done at other facilities could be obtained. These issues need additional consideration before binding policies are established.

Inflammatory pseudotumor of the liver is a localized mass consisting of a fibrous stroma and chronic inflammatory infiltrate without anaplasia. Diagnosis of this rare disease is important to avoid surgery. The purpose of this study was to determine if CT is useful in the diagnosis of this lesion.

CT scans of nine patients with a proved diagnosis of inflammatory pseudotumor of the liver were reviewed. Diagnosis was made by the surgical resection in three patients and by percutaneous biopsy in six patients. Six patients had symptoms and laboratory data suggesting active inflammation caused by the pseudotumor. The remaining three patients were asymptomatic. CT scans were performed with IV administration of the contrast material; scans were obtained in the portal venous and delayed phases in six patients and in the delayed phase in three patients. CT scans were analyzed for the number and size of the hepatic masses, and the degree and pattern of contrast enhancement on portal venous phase and delayed-phase images.

Eight patients had a solitary hepatic mass, and one patient had two masses on the CT scan. The average size of the masses in the symptomatic patients (8.3 cm) was larger than that in the asymptomatic group (3.6 cm). CT scans in the portal venous phase showed a variable degree of contrast enhancement (seven masses). At least a part of seven masses, six of which were in symptomatic patients, showed greater contrast enhancement on delayed-phase CT scans than on the normal liver parenchyma. No constant pattern of enhancement was observed on delayed-phase CT scans in asymptomatic patients.

Inflammatory pseudotumor of the liver should be included in a differential diagnosis in patients with a hepatic mass on a CT scan, especially when patients are symptomatic and the mass is fairly large and solitary showing contrast enhancement greater than that of liver parenchyma on delayed-phase CT scans. Percutaneous biopsy should be performed to obtain a histologic confirmation.

Wedge-shaped areas of increased signal intensity surrounding focal hepatic lesions on T2-weighted images have been described as an occasional finding in patients with hepatic metastases. We reviewed MR images of patients with benign and malignant focal hepatic lesions to determine the prevalence and diagnostic significance of this finding, and in particular to determine if these wedge-shaped areas are characteristic of cancer.

One hundred twenty-one patients with focal hepatic lesions (65 patients with metastases, 14 patients with hepatocellular carcinoma, and 42 patients with benign conditions) underwent MR imaging of the liver at 1.0 T. Axial spin-echo T1-weighted and spin-echo and turbo spin-echo T2-weighted MR images were obtained. The MR images were retrospectively evaluated: the number and size of the lesions and the presence of wedge-shaped areas surrounding the lesions were recorded; in patients with cancer, follow-up MR images and a clinical course were used to study the evolution of the lesions and the adjacent wedge-shaped areas. The final diagnosis in patients with cancer was proved by clinical course (54 patients) or fine-needle aspiration biopsy (25 patients); in patients with benign lesions, the findings on various other imaging techniques were accepted as conclusive.

The wedge-shaped areas were noted on T2-weighted images in 16 (25%) of 65 patients with metastases, in 3 (21%) of 14 patients with hepatocellular carcinoma, and in 5 (12%) of 42 patients with benign lesions. In 11 of the 16 patients with metastases and wedge-shaped areas of increased signal intensity, a metastatic lesion within the wedge-shaped area was detectable; in the remaining five patients, a metastatic lesion appeared during the subsequent clinical course. In some instances, the wedge-shaped area obscured the lesion. In patients with benign conditions, the wedge-shaped area was frequently seen in association with hemangiomas (4 of 13, 31%).

Our results show that wedge-shaped areas of increased signal intensity surrounding lesions on T2-weighted MR images are a common finding in patients with focal hepatic lesions. The wedge-shaped areas cannot be considered pathognomonic of malignant lesions because they are also seen in patients with benign conditions (hemangioma, abscess). However, the appearance of these wedge-shaped areas in a patient with cancer should suggest a metastasis. The possibility that hepatic lesions could be obscured by these wedge-shaped areas also should be considered when response to treatment is being evaluated.

The purpose of this study was to evaluate the clinical usefulness of a galactose-based, IV sonographic contrast agent for assessing tumor vascularity and diagnosing hepatocellular carcinoma.

We used color Doppler sonography with the sonographic contrast agent to examine 22 patients with 26 hepatic nodules (18 hepatocellular carcinomas, four hemangiomas, two adenomatous hyperplasias, and two metastatic tumors). In all 26 lesions, intratumoral arterial flow signals were examined before and after IV injection of the sonographic contrast agent at three concentrations (200, 300, and 400 mg/ml), and the findings on color Doppler sonograms of each lesion were correlated with angiographic findings.

Conventional color Doppler sonograms showed flow in nine hepatocellular carcinomas (50%) and one hemangioma (25%). When the contrast agent was used, color Doppler sonograms showed intratumoral arterial flow in 11 hepatocellular carcinomas (61%) and one hemangioma (25%) at a concentration of 200 mg/ml, in 14 hepatocellular carcinomas (78%) and 1 hemangioma (25%) at 300 mg/ml, and in 15 hepatocellular carcinomas (83%) and two hemangiomas (50%) at 400 mg/ml. The detectability of intratumoral arterial flow was improved by the contrast agent, especially in hepatocellular carcinomas smaller than 30 mm in diameter. Angiography revealed neovascularization or staining in 15 hepatocellular carcinomas, four hemangiomas, and none of the adenomatous hyperplasias or metastatic tumors. Among 15 angiographically hypervascular hepatocellular carcinomas, the detection rate of intratumoral arterial flow with contrast-enhanced color Doppler sonography was 73% at 200 mg/ml, 93% at 300 mg/ml, and 100% at 400 mg/ml. No intratumoral Doppler signals were depicted with the use of contrast agent in any angiographically undetected tumors.

Preliminary findings on contrast-enhanced color Doppler sonograms correlate well with angiographic findings for evaluating tumor vascularity. This noninvasive technique may be useful in diagnosing hypervascular hepatocellular carcinomas.

Hereditary hemorrhagic telangiectasia (HHT) or Osler-Weber-Rendu disease is an autosomal dominant disorder characterized by telangiectases and arteriovenous malformations (AVM) of skin, mucosa, and potentially every organ. The reported prevalence of the disorder is 1/100,000-2/100,000; hepatic involvement occurs in 8-31% of cases (almost always defined by clinical criteria). Hepatic involvement is shown by examination of biopsy and necropsy specimens [1] and by angiography [2]. More recently, hepatic vascular malformations have been detected by using sonography [3], Doppler sonography [4, 5], CT [6], and MR imaging [7]. In almost all the cases described, the malformations were detected in a late stage of hepatic vascular derangement. This pictorial essay illustrates the broad spectrum of abnormalities of hepatic vessels and collaterals in HHT that are detectable by imaging techniques even in the early or clinically silent stages of the diseases.

We studied the value of sonography in determining the site and cause of colonic obstruction.

Sonographic findings in 26 patients with known (n = 21) or suspected (n = 5) colonic obstruction on the basis of clinical findings and abdominal radiographs were correlated with radiologic and surgical findings. Colonic obstruction was proved in all with findings from surgery (n = 18) or from barium enema and CT (n = 8). Causes of the obstructions included colorectal cancer (n = 13), ileocolic intussusception (n = 11), transverse colonic adhesion (n = 1), and sigmoid volvulus (n = 1). Sonographic criterion of obstruction was the demonstration of a continuous distension of colonic loop with an abrupt transition to an empty distal colon. The value of sonography in terms of indicating the level and cause of colonic obstruction was evaluated.

The location of colonic obstruction was established by using sonography in 22 (85%) of 26 cases, and the cause of obstruction was identified in 21 (81%) cases. Sonography depicted a mass (n = 5) or a segmental wall thickening (n = 5) in cases of colon cancer, and a target or doughnut sign in cases of ileocecal intussusception (n = 11). Sonography failed to depict the cause of obstruction in three cases of colon cancer and one case each of adhesion and sigmoid volvulus.

Our experience suggests that sonography is useful for examining patients with colonic obstruction to determine the level and cause of the obstruction.

We evaluated a variety of internal architectural features in ruptured and nonruptured abdominal aortic aneurysms to determine whether any features are associated more frequently with ruptured abdominal aortic aneurysms. These features may be useful in identifying subtle ruptures when no obvious retroperitoneal hematoma is present and may be helpful in predicting unstable aneurysms at risk for rupture.

The CT scans of 52 patients with ruptured abdominal aortic aneurysms were reviewed and compared with those of 56 patients with asymptomatic nonruptured aneurysms exceeding 4.5 cm in diameter. All aneurysms were evaluated for size, rim calcification, thrombus amount, thrombus calcification, and lumen irregularity. In addition, four different thrombus patterns were identified and evaluated, including homogeneous, diffusely heterogeneous, periluminal halo, and crescent patterns. Statistical comparisons were adjusted for differences in size between the two groups.

Age, gender, and aneurysm length were not statistically different between the two groups. A larger diameter was found in the ruptured aneurysm group: 7.4 (anteroposterior) x 7.9 (transverse) cm versus 5.9 x 6.1 cm (p = .00001). More thrombus surrounded the nonruptured aneurysms (p = .014). Thrombus calcification was seen in 25% (14/56) of the control group and in 13% (7/52) of the rupture group (p = .01). Two thrombus patterns, homogeneous and periluminal halo, were encountered with similar frequencies in both groups. The diffusely heterogeneous pattern was seen more in the control group. A crescent of increased attenuation was encountered only in patients with ruptured aneurysms, at an incidence of 21% (11/52) (p = .0005). Thick and thin wall calcifications were seen in both groups, but a focal discontinuity in circumferential calcification was seen only in association with ruptured aneurysms, at an incidence of 8% (4/52) (p = .008). There was no significant difference in the number of patients whose patent lumen was irregular.

In our series, detection of a high-attenuation crescent or focal gap of otherwise circumferential wall calcification is associated with aneurysm rupture. The homogeneous, diffusely heterogeneous, and periluminal halo patterns are not specifically associated with aortic rupture. There were no significant differences in the amount of wall calcification or frequency of lumenal irregularity between patients with ruptured and those with nonruptured aneurysms.

We performed this study to compare contrast-enhanced fast multiplanar spoiled gradient-recalled (SPGR) MR imaging with unenhanced spin-echo imaging for the detection of benign and malignant peritoneal abnormalities.

We retrospectively reviewed abdominal MR images of 34 patients with proved peritoneal abnormalities, including 21 patients with malignant and 13 patients with benign peritoneal disease. Six additional patients had false-positive interpretations of their MR examinations because of diaphragmatic thickening and enhancement. Unenhanced T1-weighted images, fast spin-echo T2-weighted images, and breath-hold fast multiplanar SPGR images obtained immediately and 10-15 min after IV injection of gadopentetate dimeglumine were available. Images were reviewed for evidence of peritoneal disease by two observers who had no knowledge of the clinical findings. The four types of images were separately evaluated for peritoneal thickening, masses, and abnormal enhancement. The anatomic location, thickness, and pattern of thickening of the peritoneum were noted. Results were correlated with surgical findings or results of percutaneous biopsy.

Immediate and delayed contrast-enhanced fast multiplanar SPGR images showed malignant peritoneal tumor in all 21 cases, compared with T1-weighted images, which showed tumor in 11 cases (p < .01), and fast spin-echo T2-weighted images, which showed tumor in 13 cases (p < .01). Enhancement of the peritoneal tumor increased the tumor's conspicuity, particularly on delayed fast multiplanar SPGR images, which were selected as the best sequence for detection of peritoneal tumor in 18 of the 21 cases. The MR findings varied widely from moderately and regularly thickened, enhancing peritoneum to very thick (> 1 cm) and lobular or masslike peritoneal thickening with enhancement. Delayed fast multiplanar SPGR images showed evidence of benign peritoneal disease in all 13 cases compared with immediate fast multiplanar SPGR images, which showed disease in 12 cases (p > .05); T1-weighted images, which showed disease in four cases (p < .05); and fast spin-echo T2-weighted images, which showed disease in only one case (p < .01). In the benign forms of peritoneal disease, the peritoneum was often regular in contour and tended to be thinner than in the cases of malignant peritoneal tumor, in which the peritoneum was generally thicker and more lobular or masslike. However, significant overlap was seen between the MR findings in benign and malignant forms of peritoneal disease.

Detection of peritoneal abnormalities with MR imaging can be improved by using contrast-enhanced fast multiplanar SPGR imaging. The findings of peritoneal thickening and enhancement are best depicted on delayed images.

The purpose of this study was to evaluate the clinical efficacy of transvaginal sonographically guided aspiration and drainage of pelvic fluid collections.

Forty patients underwent transvaginal sonographically guided aspiration of a possible pelvic abscess (41 pelvic collections). In patients with clinical findings highly suggestive of infection, both purulent and nonpurulent collections were immediately drained via a catheter. In patients with clinical findings moderately suggestive of infection, nonpurulent collections were completely removed by aspiration and the aspirates were cultured; however, purulent collections were immediately drained via a catheter.

All collections were successfully accessed by transvaginal sonography. For 27 of the 41 collections, the aspirate was purulent (18 collections) or the patient's clinical findings were highly suggestive of infection (nine collections) and catheter drainage was performed. Seventeen of the 27 collections completely resolved and surgery was not required. Four of the 27 collections were in patients who had surgery for reasons other than persistent infected collection. For six of the 27 collections, catheter treatment was not successful and surgery was required. The overall success rate of catheter drainage was 78%. In the remaining 14 of the 41 collections, the aspirate was serous or serosanguineous, and the patient's clinical findings were moderately suggestive of infection. Cultures of aspirates of seven collections were positive for microorganisms. Eleven collections were successfully treated with antibiotics or no therapy was required (based on culture results); for three, surgery was required. Two complications occurred: one vaginal fistula after catheter drainage and one disruption of vaginal sutures after aspiration.

Transvaginal sonographically guided drainage is effective treatment of pelvic abscess, being either completely curative or temporizing in 78% of patients. Catheter treatment was unsuccessful and surgery was necessary in 22% of patients. For nonpurulent collections, catheter drainage is indicated only when clinical findings are highly suggestive of infection.

The purpose of this study was to determine whether pulsed Doppler sonography can be used to distinguish between benign and malignant adnexal masses on the basis of pulsatility index.

In an 18-month period, all patients in whom an adnexal mass was detected at sonography had further evaluation of the mass by color and pulsed Doppler sonography. Ninety-nine patients with 102 masses that were surgically removed were included in the study. The pulsatility indexes were calculated from the reproducible spectral waveforms generated from flow centrally or peripherally within or immediately adjacent to the mass. Each lesion was categorized on the basis of its gray scale morphologic features as typically benign or indeterminate/malignant in appearance.

Of the 102 adnexal masses, 89 were benign and 13 were malignant. In seven of the 89 benign lesions, no flow could be detected, and these were excluded from analysis. Of the remaining 82 benign lesions, 65 showed pulsatility indexes consistently equal to or greater than 1.0, and 17 showed pulsatility indexes of less than 1.0. Ten of the 13 malignant lesions had pulsatility indexes consistently less than 1.0, and three primary malignant tumors had their lowest pulsatility indexes ranging between 1.1 and 1.8. Sixty-five of the 68 masses with pulsatility indexes equal to or greater than 1.0 were benign, for a positive predictive value of 96% for benign disease. Ten of the 27 masses with pulsatility indexes of less than 1.0 were malignant, for a positive predictive value of 37% for malignant disease. Forty-five masses were detected in perimenopausal and postmenopausal patients. In this group, the pulsatility index had a positive predictive value of 88% for benign disease and 47% for malignant disease. In this study, 45 of 49 masses that had a typically benign sonographic appearance had pulsatility indexes equal to or greater than 1.0. All 49 masses had benign histology. In the remaining 46 masses with an indeterminate/malignant sonographic appearance, 20 of 23 with pulsatility indexes equal to or greater than 1.0 were benign, and 10 of 23 with pulsatility indexes of less than 1.0 were malignant.

Our results show a high positive predictive value of high-impedance flow in benign adnexal disease and a predominance of low-impedance flow in malignant adnexal disease. However, the pulsatility indexes showed considerable overlap between benign and malignant lesions, indicating that Doppler sonography has severe limitations in the differentiation of benign from malignant adnexal disease on the basis of low-impedance flow (pulsatility index < 1.0).

MR imaging has revolutionized the diagnostic evaluation of musculoskeletal disorders, becoming the study of choice for many indications. Without the use of ionizing radiation, MR imaging can safely produce high-contrast, high-resolution cross-sectional images of the body in virtually any imaging plane. To maximize the diagnostic information of musculoskeletal MR studies, the radiologist must consistently produce the highest quality images possible, which in turn depends on a complex interaction among multiple technical factors. We present our perspectives on some of these factors, stressing the impact each has for specific indications. The material presented herein should be viewed as guidelines for improving image quality and diagnostic yield. However, individual systems and applications vary considerably, and the radiologist must "fine-tune" protocols to reflect the available resources, patient population, and preferences of the referring clinicians.

Imaging studies of patients with rotational facet injuries of the cervical spine were retrospectively reviewed to determine the prevalence and pattern of associated fractures, to correlate injury pattern with recommended surgical stabilization, and to assess neurologic outcome.

Radiographs and CT scans obtained for 40 consecutive patients with rotational facet injuries of the cervical spine during a 70-month period were retrospectively reviewed to determine injury level, presence, and orientation of facet fractures, and concurrent nonfacet injuries. Imaging findings were reviewed to assess the likelihood of instability and to determine the most appropriate stabilization requirement. Medical records were reviewed to ascertain mechanism of injury, initial neurologic deficit, and surgical findings.

Among the 40 patients with cervical rotational facet injuries, 11 (27%) had pure unilateral facet dislocation or subluxation without associated fractures, and 29 (73%) had concurrent facet fractures involving the inferior facet of the rotated vertebra (n = 13), the superior facet of the subjacent vertebra (n = 9), or both (n = 7). Injury of the rotated vertebra was unilateral in 22 patients but bilateral in 18 patients. Facet fractures frequently extended into the ipsilateral lamina or articular pillar or both. An avulsion fracture from the posteroinferior aspect of the rotated vertebral body, indicating disk disruption, occurred in 10 patients (25%), and seven patients (17%) had complete isolation of an articular pillar. Facet fractures were confirmed for 27 patients who underwent surgical stabilization. Neurologic deficits developed in 29 (73%) of the 40 patients and included radiculopathy in 11 patients and cord syndromes in 18 patients. Pure dislocation without a facet fracture was more likely to lead to a cord syndrome (p = .006).

Cervical rotational facet injuries are often accompanied by facet fractures and bilateral damage of the rotated vertebra. These injuries contribute to rotational instability and require specific internal fixation based on a precise delineation of all injuries. Facet dislocations without fractures have a significantly higher association with cord syndromes than do rotational facet injuries with fractures. CT, particularly with parasagittal reformations, is valuable in identifying all injuries of the rotated and subjacent vertebrae.

Selecting the appropriate surgical procedure for treatment of patients with osteosarcoma requires accurate preoperative evaluation of tumor extent. Establishing the presence or absence of joint involvement is particularly important. Accordingly, we studied the efficacy of MR imaging for detecting joint involvement in 46 patients with osteosarcoma around joint spaces.

Preoperative MR examinations were performed in 46 consecutive patients with osteosarcoma whose tumors were located around the knee (n = 33), the hip (n = 8), or the shoulder (n = 5). T2-weighted and unenhanced and contrast-enhanced T1-weighted spin-echo MR images were obtained for all patients. We assessed the presence or absence of tumor invasion of the intracapsular-intrasynovial joint space, either by disruption of the joint capsule or by intraarticular destruction of the cortical bone and articular cartilage or the intracapsular-extrasynovial cruciate ligaments of the knee. All patients subsequently had surgery. The MR findings were correlated with findings from macroscopic and microscopic pathologic examinations.

All 10 patients who subsequently proved to have tumor involvement of the joint were correctly identified (sensitivity, 100%). The tumor involved the knee joint in seven patients, the hip joint in two, and the shoulder joint in one. However, the MR diagnosis was false-positive in another 11 patients who did not have joint involvement at surgery (specificity, 69%). In the knee, MR imaging was more accurate in identifying tumor extension to the cruciate ligaments than to the intrasynovial joint space. Post-contrast T1-weighted images were most useful in detecting joint involvement.

MR imaging is highly sensitive for detecting joint invasion of osteosarcoma. However, false-positive diagnoses may lead to overstaging of tumor and result in unnecessarily radical surgical procedures.

The purpose of this in vitro study was to determine the value of spiral CT for detecting displacement of fractures of the tibial plateau. The exact amount of inferior plateau depression, if any, is a primary criterion for deciding between surgical management and conservative management.

An artificial fracture was produced in a cadaveric tibial plateau. Inferior displacements of 0, 1, 3, and 5 mm were created at the fracture. Spiral CT scans of each displacement were obtained with table speeds of 2, 3, and 5 mm/sec. Section collimation equaled table increment. Coronal image reconstructions were produced by use of standard scanner software. The images were interpreted by six musculoskeletal radiologists.

Interpretation accuracy was greatest with coronal images created from the 2 mm/sec scans. For distinguishing 5-mm fracture displacements, the average diagnostic sensitivity and specificity were 96% and 93%, respectively; when a 2-mm depression was used as the criterion for clinical significance, the sensitivity was 100% and the specificity was 69%.

When minimal table increment and collimation are used, spiral CT can detect clinically important inferior depressions of tibial plateau fractures. On the basis of the results of this study, when spiral CT is used for tibial plateau fracture assessment, we recommend 2-mm section collimation, 2-mm table speed, and reconstruction of images at 1-mm increments.

The normal anatomy of the long head of the biceps tendon of the shoulder has been described in detail [1]. Descriptions of different pathologic processes affecting this structure also have been published [1-3] but have been incomplete, showing only a limited variety of abnormalities. In this article, abnormalities of the long head of the biceps tendon seen on MR images are illustrated in greater variety and detail. Recognizing abnormalities of the biceps tendon is important because they are a common source of shoulder pain both alone and in combination with abnormalities of the rotator cuff, labrum, and other structures. Because incomplete diagnosis can lead to treatment failure, it is important to recognize less common imaging manifestations of common entities.

The use of color Doppler sonography to diagnose scrotal disorders in children has been hampered by the small size of the vessels and the slow blood flow compared with those in adults. Spectral analysis is the best means available of confirming the vascular origin of questionable color-flow signals arising from testes of small volume. The range of normal and abnormal Doppler sonographic arterial waveforms arising from the testis in boys, as distinct from those in adults, has never been described. The purpose of our study was to establish the normal range of testicular arterial impedance, measured as resistive index (RI), in both prepubertal and pubertal/post pubertal boys.

Bilateral scrotal color Doppler sonography was performed in 33 healthy boys aged 3 days to 17.5 years. The mean RI in the parenchymal artery was measured in 44 testes and plotted against testicular volume. Chi-square statistics were used to test differences in mean RI, testicular volume, and age between testes with volumes of 4 cm3 or less and testes with volumes greater than 4 cm3.

In testes with volumes of 4 cm3 or less, the mean RI ranged from 0.39 to 1.00 (mean, 0.87), and in testes with volumes greater than 4 cm3, the RI ranged from 0.43 to 0.75 (mean, 0.57). In 20 of 30 testes with volumes of 4 cm3 or less, the RI of parenchymal arteries was equal to 1.00 (i.e., diastolic flow was undetectable). Mean RI, testicular volume, and age were significantly different between testes with volumes of 4 cm3 or less and testes with volumes greater than 4 cm3 (p < .001 for all variables).

Mean testicular RI in our samples of pubertal and postpubertal boys is decreased compared with the mean RI in prepubertal boys. Diastolic arterial flow may not be detectable in normal testes with volumes of 4 cm3 or less. Despite the existence of a statistically significant difference in mean RI between prepubertal and pubertal/postpubertal testes, substantial variability exists within each group, particularly among the prepubertal testes. Norms of testicular flow previously established for adults are therefore not routinely applicable to prepubertal boys with testicular volumes less than 4 cm3. However, RI values in normal pubertal and postpubertal boys where testicular volumes exceed 4 cm3 are comparable to those previously described in adults.

The purpose of this study was to determine the sensitivity and specificity of CT in diagnosing bowel rupture in children after blunt trauma and to compare CT findings in children with bowel rupture with those in children with the hypoperfusion complex.

Twenty-one (1%) of 1488 children who had contrast-enhanced CT of the abdomen after blunt trauma had a bowel rupture subsequently verified at surgery or autopsy. Thirty-three additional children had a characteristic hypoperfusion complex at CT. The CT scans in all 1488 children were prospectively evaluated for the following findings: peritoneal fluid, extraluminal air, bowel wall enhancement, bowel wall thickening, and bowel dilatation.

The most common CT findings in children with bowel rupture were peritoneal fluid (14, 67%) and bowel wall enhancement (13, 62%). One or more of the five studied CT findings were noted in 20 of the 21 children with bowel rupture and in 64 of the 1467 children without bowel rupture (sensitivity 95%, specificity 96%). Thirty-three children who had one or more of the CT findings and did not have bowel rupture had the hypoperfusion complex.

Our results show that CT is accurate in the diagnosis of bowel rupture after blunt trauma in children. The most common findings are peritoneal fluid and bowel wall enhancement. CT findings in children with bowel rupture may overlap with those in children with the hypoperfusion complex.

The purpose of this study was to determine by MR imaging the prevalence and types of dysraphic abnormalities of the spinal cord (i.e., myelodysplasias) associated with urogenital and anorectal malformations of childhood.

Since 1987, 92 patients with imperforate anus complex, cloacal malformation, and cloacal exstrophy have had MR imaging as a screening examination for occult dysraphic myelodysplasia. The prevalence and types of myelodysplasia were determined for each group.

The prevalence of dysraphic myelodysplasia in each group of children was 17% (1/6) for low imperforate anus (ectopic anus), 34% (11/32) for high imperforate anus (with fistulization), 46% (19/41) for cloacal malformation, and 100% (13/13) for cloacal exstrophy. The most common abnormalities in each group were tethered cord with intradural or filar lipoma in imperforate anus; low-placed or dysplastic conus medullaris and tethered cord with lipoma or myelolipoma in cloacal malformation; and lipomyelocele, lipomyelomeningocele, or lipomyelocystocele in cloacal exstrophy.

Our results show that the prevalence of myelodysplasia as seen on MR imaging is high in patients with urogenital and anorectal anomalies.

The purpose of this prospective study was to compare MR angiography of the carotid artery from the aortic arch through the circle of Willis using maximum-intensity projection (MIP) and multiplanar reformation (MPR) images with intraarterial angiography in the depiction of extracranial carotid atherosclerosis.

The carotid arteries in 20 patients were studied with MR and intraarterial angiography. MR angiography included two-dimensional (2D) time-of-flight (TOF) sequences from the aortic arch through the skull base and three-dimensional (3D) TOF sequences centered at the carotid bifurcation and multiple overlapping thin slab acquisition (MOTSA) from the skull base to above the circle of Willis. Targeted MIP images of the 2D and 3D TOF MR angiograms through each carotid bifurcation were obtained. Last, MPR images of the 3D TOF MR angiograms at the obliquity that showed the greatest stenosis were obtained. All studies were reviewed in a double-blinded fashion by two neuroradiologists. Caliper measurements of MR angiograms and intraarterial angiograms were made by using North American Symptomatic Carotid Endarterectomy Trial criteria. Global MIPs of the aortic arch and common carotid arteries from the 2D TOF MR angiograms and targeted MIPs of the intracranial carotid arteries from the MOTSA MR angiograms were compared with the intraarterial angiogram and graded as normal, mild, moderate, severely stenotic, or occluded.

MPR of the 3D TOF MR angiograms was highly correlated with intraarterial angiograms for both observers 1 and 2 (0.94/p < .001, 0.96/p < .001 [Pearson correlation/p value]). No statistically significant difference between 3D TOF MPR and intraarterial angiography was seen with a paired t-test. With an alpha = 0.05 (5% probability of type 1 error), the power to detect a difference as small as +/- 5% stenosis between 3D TOF MPR and intraarterial angiogram was 80% for observer 1 and 90% for observer 2. Although both MIPs of the 2D and 3D TOF MR angiograms showed high Pearson correlation coefficients (0.83, 0.90) with intraarterial angiography, the paired t-test revealed a statistically significant difference in the estimation of carotid stenosis. Both observers thought the global MIPs of the 2D TOF MR angiogram allowed good to excellent visualization of the common carotid arteries. The aortic arch was seen in 70% of patients; most of the missed cases occurred early in our experience, when the 2D axial images were not placed sufficiently inferior to include the arch. No stenosis of the great vessel origins was seen in this study. All four stenoses of the intracranial internal carotid artery identified with intraarterial angiography were seen with the MOTSA MR angiogram but with a tendency to overestimate stenosis. Only one carotid siphon was thought to show severe stenosis on the MR angiogram. Intraarterial angiography showed a 50% stenosis.

It is possible to image the entire carotid artery from the aortic arch through the circle of Willis with MR angiography in a clinically acceptable time. MPR of the 3D TOF MR angiogram reliably shows the percentage of carotid stenosis with no statistically significant difference compared with intraarterial angiography. The role of MR angiography in showing lesions in the circle of Willis or the aortic arch is promising, but the limited number of tandem lesions in this study makes it difficult to draw any conclusions.

The basal gray matter in the cerebrum is supplied by different anatomical groups of perforating arteries, including striate, thalamic, and choroidal arteries. In this pictorial essay we illustrate multiplanar MR imaging of infarctions in those regions and correlate their appearances with cerebral angiographic findings of the same patients when available. We also correlate the extent and location of infarctions on multiplanar MR images with the anatomical distributions of perforating arteries seen on microangiograms of unrelated cadavers. Information conveyed from this correlation will increase understanding of patterns of basal cerebral infarction shown on cross-sectional imaging.

The purpose of this study was to assess the value of measuring parathormone levels in percutaneous needle aspirates of suspicious cervical lesions in patients with hyperparathyroidism to confirm whether the lesion represents abnormal parathyroid tissue.

The study group consisted of 66 patients with hyperparathyroidism in whom 80 cervical lesions were aspirated and levels of parathormone in the aspirates were measured. CT guidance was used for two patients and sonographic guidance for the remainder. The lesions selected for aspiration were demonstrated on either sonography or CT and had either an unusual position (separate from the thyroid gland or were intrathyroidal) or configuration (irregular shape or atypical heterogeneous sonographic texture). In 15 patients, an indeterminate, posteriorly located intrathyroidal mass was detected and felt most likely to represent a thyroid nodule by sonographic criteria. These masses were aspirated to rule out atypical parathyroid adenomas. In patients who had been previously explored for hyperparathyroidism and presented with persistent or recurrent hypercalcemia, all indeterminate, cervical, potentially parathyroid masses were aspirated for parathormone determination. The level of parathormone in each aspirate was measured by using an immunoradiometric assay.

Levels of parathormone were increased in the aspirates in 37 of the 45 patients in whom sonography showed classic lesions suggestive of parathyroid adenoma. This included the 25 patients who had previously undergone exploratory surgery. At surgery, all 37 had parathyroid adenomas in the indicated locations, for a specificity of 100%. In six patients, the results of the aspiration were false-negative. Parathyroid adenomas were suspected on sonograms and confirmed at surgery, but no parathormone was detected in the aspirate. Results of aspiration of indeterminate lesions were true-negative in two patients who had both characteristic and indeterminate lesions on sonography and in the 15 patients who had indeterminate lesions that were felt to be of thyroid origin. Aspirates contained no parathormone, and surgical findings confirmed the lesions were not of parathyroid origin.

Our results show that increased levels of parathormone in percutaneous needle aspirates of cervical masses in patients with hyperparathyroidism confirm the mass is a parathyroid adenoma. Although absence or low levels of parathormone in the aspirates usually excludes a parathyroid adenoma, this is not absolute as sometimes the needle may miss the mass, rendering the parathormone value invalid (false-negative).

The purpose of this project was to create an approach for global radiology multimedia publishing using the internet that would address the two largest problems facing radiology multimedia publishers today: the high percentage of radiologists who are computer novices and the variety of personal computers (Macintosh, Microsoft Windows/IBM-PC, X-Windows, Amiga) whose software is incompatible.

We developed a client/server approach to multimedia publishing, the networked multimedia textbook, that has a simple booklike user interface to facilitate use by computer novices. Once created, a networked multimedia textbook can be viewed on all current popular personal computers. The networked multimedia textbook is based on the internet, World-Wide Web, Mosaic, and Wide Area Information Servers software technologies, all of which are in the public domain.

We created six radiology networked multimedia textbooks.

This networked multimedia textbook approach for the global distribution of multimedia radiology information brings the benefits of multimedia publishing on the Internet to radiologists today.

Recent advances in storage technology have made possible the archiving of tremendous amounts of text and image information within a picture archiving and communication system (PACS). However, a radiologist's access to this information typically has been limited to viewing workstations designed primarily to support clinical activities. Unfortunately, these workstations often overlook the benefits of PACS in teaching and research applications, which are of significant importance in an academic institution. To support such activities at our own institution, we have included two major objectives in our second-generation PACS development: (1) to provide access to text and image information archived within our PACS in an environment that is easily accessible to and comfortable for our radiologists, namely, their own Macintosh (Apple Computer, Cupertino, CA) personal computers; and (2) to provide this information in standard Macintosh formats, so that tools with which radiologists are already familiar can be used in frequently performed teaching activities--the production of slides and prints, the maintenance of personal teaching and research files, and specialized image analysis and processing.

The main objectives of the study were as follows: first to study the nature and extent of radiologists' involvement in and their attitudes toward quality assessment (QA) and continuous quality improvement (CQI)/total quality management (TQM) in hospitals and in offices; and second, to ascertain whether differences in size, type, and location among hospitals and nonhospital radiology offices affect the QA and CQI/TQM activities of radiologists. We analyzed data from a national survey conducted by the American College of Radiology (ACR) in 1993.

Questionnaires about QA and CQI/TQM activities and attitudes were mailed to 216 hospital-affiliated diagnostic radiology group practices using a sample selected from the ACR master list of radiology practices in the United States. The response rate was 90%. A stratified random sample ensured representation of different geographic regions, various group sizes, and both academic and nonacademic groups. Responses were weighted so that our data show what answers about hospitals would have been if (i) the survey had been answered by all hospital radiology departments in the United States (except for those few staffed by solo practitioners or nonradiologists) and (ii) our questions about nonhospital offices had been answered by all radiology groups in the United States (except those few having no hospital activity).

The majority (86%) of hospital radiology departments report having a program to monitor and evaluate physicians' performances. Fifty-one percent collect incorrect diagnoses by specific radiologist. Twenty-eight percent collect some of their QA data through computerized information systems. We found some statistically significant differences by hospital size and location, with larger hospitals and urban hospitals being more likely to engage in some QA activities. Multivariate analyses, once controlled for hospital size and location, found no significant differences in QA activity between university and community hospitals or between hospitals with and without a residency program. QA and CQI programs were less common in offices than in hospitals. With the exception of mammographic interpretations, most practices did not monitor and evaluate physicians' performances in the office setting. Respondents representing 58% of hospital radiology departments thought that QA and CQI contributed to improvement in patient care. Only 19% of radiology practices answered that CQI has been of cost benefit to their organization.

Most radiology practices engage in a variety of QA and CQI activities in hospitals. However, this is less true in offices, in which radiologists have more discretion, and radiologists remain skeptical about the usefulness of CQI.