Rüdiger Fründ

Hepatic encephalopathy is a common problem in cirrhosis. The pathogenesis of this complication of advanced liver disease still remains unclear. Magnetic resonance spectroscopy was used to assess prospectively cerebral metabolism in 51 patients with histologically proven cirrhosis (Child–Pugh classes A, B, and C, 18, 18, and 15, respectively) and 36 healthy volunteers. According to the results of psychometric tests, overt hepatic encephalopathy, subclinical encephalopathy, and no encephalopathy were found in 14, 21, and 16 patients, respectively. Myoinositol/creatine ratios in gray (.36 ± .17) and white (.35 ± .22) matter voxel were reduced significantly ( P < .0001) in cirrhotic patients compared with healthy volunteers (gray matter, .51 ± .11; white matter, .64 ± .16). In addition, patients showed a significant reduction ( P = .024) in white matter choline/creatine ratio (.77 ± .27) compared with controls (.92 ± .25), and glutamine/glutamate level was elevated in cirrhotic patients compared with controls (gray matter, P < .0001; white matter, P = .036). Changes in cerebral myoinositol and glutamine/glutamate levels correlated significantly with the severity of hepatic encephalopathy ( P < .0001). However, these metabolic alterations were also detected in patients without hepatic encephalopathy (normal psychometric test results). N –acetyl aspartate/creatine ratios did not differ between patients and controls. Magnetic resonance imaging detected bright basal ganglia in 37 patients, which correlated significantly with portal–systemic shunting and elevation of glutamine/glutamate, but not with the degree of hepatic encephalopathy. In conclusion, magnetic resonance imaging and spectroscopy showed that alterations of cerebral metabolism are common in patients with cirrhosis, even without evidence of clinical or subclinical hepatic encephalopathy.

RATIONALE AND OBJECTIVES: The authors compared a solid-state amorphous silicon (a-Si) detector and screen-film radiography (SFR) with regard to the detection of simulated pulmonary lesions. Evaluation of the impact of a dose reduction of 50% with this digital flat-panel detector was of special interest. METHODS: A self-scanning flat-panel detector, based on a-Si technology with 143 x 143 microm pixel size, 1 k x 1 k matrix and 12-bit digital output was used. An asymmetric state-of-the-art screen-film system was compared with a-Si images taken at the same dose as SFR-images and at a dose reduced by 50%. An anthropomorphic chest phantom was superimposed by templates containing nodules, linear structures, reticular, and micronodular opacities in a random distribution. Receiver operating characteristic analysis was performed for 23,040 observations made by four independent observers. Student's t test (95% confidence-level) was used for statistical analysis. RESULTS: Receiver operating characteristic analysis showed that a-Si images taken at the same dose as SFR-images were significantly superior to SFR with respect to the detectability of lines (P = 0.01) and micronodular opacities (P < 0.01). For the other objects and the a-Si images taken at a reduced dose, it yielded no statistically significant differences between both imaging modalities. CONCLUSIONS: The results of this phantom study indicate that a-Si detector technology holds promise in terms of dose reduction in chest radiography without loss of diagnostic accuracy compared with SFR.

OBJECTIVE: The purpose of this study was to evaluate the effectiveness of a large-area, flat-panel X-ray detector for performing routine chest radiography at two different detector doses. MATERIALS AND METHODS: The chest radiographs of 50 patients (age range, 16-79 years; mean age, 57 years) were obtained at two different detector dose levels. Digital images were taken from the same patients in posteroanterior and lateral views with detector doses of 2.5 microGy and 1.8 microGy, respectively, at 125 kVp tube voltage. The cesium iodide-amorphous silicon active-matrix imager had a panel size of 43 x 43 cm, a matrix of 3000 x 3000, and a pixel pitch of 143 microm. Images were presented in a random order to three independent radiologists who were unaware of the dose level at which the images had been obtained. They subjectively rated image quality on a 4-point scale, according to six criteria (presentation of obscured lung, unobscured lung, airways, mediastinum and hilum, bony thorax, and overall impression). Statistical significance of differences was evaluated with Student's t test for paired samples (confidence level, 95%). RESULTS: Digital radiographs obtained at 2.5 and 1.8 microGy were equivalent on all quality criteria. No statistically significant differences and no tendency toward a preference for images obtained at one or the other dose level were observed. According to the registered mAs values, the average difference in patient dose was 33%. CONCLUSION: Use of flat-panel digital imagers based on the cesium iodide-amorphous silicon technique allows a considerable dose reduction during routine chest radiography without loss of image quality.