Search

This study investigates the dual-energy procedure for postoperative CT follow-up scans after endovascularly treated abdominal aortic aneurysms. The procedure is analyzed with respect to its sensitivity and specificity as well as the associated radiation exposure.51 examinations were carried out on 47 patients between February 2009 and March 2010. For each patient, a non-enhanced, an arterial and a venous scan were conducted, the latter two using the dual-energy technology. Virtual images for the non-enhanced phase were reconstructed from the data taken in the venous phase. Protocol A, the reference standard, consisted of non-enhanced images and images of the arterial and venous phase. In protocol B, standard non-enhanced images were replaced by the reconstructed virtual non-enhanced images. Protocol C consisted only of virtual non-enhanced and 80 kV images taken during the venous phase. All data was anonymized and evaluated by two independent radiologists. For protocol C, sensitivity, specificity, negative and positive predictive values were computed. The effective radiation dosage was determined for each scan.All endoleaks identified in protocol A were found using protocols B and C. For protocol C, the sensitivity and negative predictive value were 100 %, the specificity was 94.1 %, and the positive predictive value was 89.5 %. Compared to protocol A, protocol C reduces the radiation exposure by 62.45 %.A scan protocol consisting of virtual non-enhanced images as well as 80 kV images taken during the venous phase was found to be a reliable alternative method for diagnosing endoleaks, while reducing the radiation exposure by 62.45 %.

Hospital hygiene is of pivotal importance in radiology departments, where patient throughput is high and staff come into direct contact with both inpatients and outpatients. Every member of the medical and technical team should be aware of all the general and specific hygiene requirements and ensure that they are considered during the daily routine. Failure to do so on the part of just one individual can result in bacterial contamination in the department, exposing both patients and staff to the risk of infection. For the purposes of performing examinations and interventions in infectious patients, the hospital hygienist and medical hygiene officer should introduce appropriate organisational structures to ensure that an appropriate level of hygiene is guaranteed and can be adapted to each patient. This will ensure a minimal degree of disruption to work flow in the radiology department. It is of particular importance that medical and technical staff receive regular hygiene training and instruction. Such training ensures that members of staff are fully aware of the hygiene requirements, are responsibly engaged and that they each possess the expertise and confidence to deal with general and specific hygiene issues in any particular case.

The general gap in the BIRADS lexicon between lesion description and categorization leads to very different recommendations in clinical breast radiology. This is particularly true for breast MRI. The third consensus meeting of course directors in breast imaging aimed at an increase in standardization of breast MRI.Between 166 and 344 participants in the audience (A) and 9 and 13 expert panel participants (P) took part in an electronic wireless voting system. The audience consisted of 98% radiologists and 2% gynecologists (A: n=295; P: n=12: radiologists 92%/gynecologists 8%). Of all participants, 62% had more than 10 years of experience in breast imaging and only 9% had less than 3 years of experience (P: 100%10 years of experience). The day before 44, clinically relevant, though unresolved questions were formulated by the expert panel. For the evaluation a distinction was made between answers with a great majority (75%), simple majority (50-75%) and no majority (50%) as well as answers from the expert panel and answers from the audience.Of 44 questions, all but two were answered with simple or great majority.Technique, reporting and clinical use are becoming more and more accurately defined in MRI of the breast and MR-guided interventions. The third consensus meeting of this kind gained numerous answers and thus enables recommendations for didactic as well as clinical routine work.

Physical-technical quality assurance is one of the essential tasks of the National Reference Centers in the German Breast Cancer Screening Program. For this purpose the mammography units are required to transfer the measured values of the constancy tests on a daily basis and all phantom images created for this purpose on a weekly basis to the reference centers. This is a serious logistical challenge. To meet these requirements, we developed an innovative software tool.By the end of 2005, we had already developed web-based software (MammoControl) allowing the transmission of constancy test results via entry forms. For automatic analysis and transmission of the phantom images, we then introduced an extension (MammoControl DIANA). This was based on Java, Java Web Start, the NetBeans Rich Client Platform, the Pixelmed Java DICOM Toolkit and the ImageJ library.MammoControl DIANA was designed to run locally in the mammography units. This allows automated on-site image analysis. Both results and compressed images can then be transmitted to the reference center. We developed analysis modules for the daily and monthly consistency tests and additionally for a homogeneity test.The software we developed facilitates the immediate availability of measurement results, phantom images, and DICOM header data in all reference centers. This allows both targeted guidance and short response time in the case of errors. We achieved a consistent IT-based evaluation with standardized tools for the entire screening program in Germany.

For the non-subjective evaluation of physical-technical defects of digital mammography systems within the German breast screening program, an automatic rating system (RS) on the basis of the annual quality control (AQC) was developed. The aim was to design a modular algorithm which could also be adjusted for other test procedures.This algorithm enables the separate evaluation of different test parts of the AQC (PAS 1054 and EPQC). To take into account the value of different test positions (TP) in these two parts, a weighting scheme was used. This scheme relates to the scope of different TP as well as variability through different types of systems. For the creation of the final result, both scores of the different test parts must be merged together. The final result is divided into 4 categories.The evaluation of the RS was based on the AQC. Here, the tester of the system was asked to classify the system in one of the 4 categories on the basis of his expert knowledge. In 78 % of the cases, the results of the tester were reproduced by the automatic RS, in 17 % the RS was more stringent and in 5 % the tester was more stringent.The RS allows non-subjective physical-technical evaluation of mammography systems. It acts as a good guide for the tester but is not a replacement for experience. In addition, the RS makes the results of different institutions and testers more comparable. The modular algorithm of the RS is able to react to future changes in the different test parts and can be modified for other X-ray modalities.

Cystic pancreatic lesions are often discovered incidentally as an asymptomatic finding, at a rate which is increasing considerably. In recent years the understanding of such tumors has become clearly differentiated. The spectrum of relevant lesions includes in particular the intraductal papillary mucinous neoplasm (IPMN), serous cystic neoplasm (SCN) and mucinous cystic neoplasm (MCN). With certain knowledge of their histological and radiomorphological structure as well as their distribution in terms of location, age and sex, such tumors are easy to differentiate and demarcate from common pancreatic pseudocysts. This also implies the fundamental understanding of complementary endoscopic procedures such as endosonography, which enables aspiration of the content of the cyst. A number of cystic pancreatic lesions have the potential to undergo malignant transformation along the adenoma-carcinoma sequence and therefore necessitate a differentiated approach to their radiological management. This review aims to develop a broad understanding of the pathological and radiomorphological characteristics of cystic pancreatic lesions and provides advice regarding procedures, particularly with respect to incidentally detected lesions.

To determine regional and global left ventricular (LV) functional parameters and to perform segmental wall thickness (SWT) and motion (WM) analysis of dual-source CT (DSCT) with optimized temporal resolution versus MRI.30 patients with known or suspected CAD, non-obstructive HCM, DCM, ARVCM, Fallot Tetralogy, cardiac sarcoidosis and cardiac metastasis underwent DSCT and MRI. The DSCT and MR images were evaluated: end-systolic (ESV), end-diastolic LV (EDV) volumes, stroke volume (SV), ejection fraction (EF), and myocardial mass (MM) as well as LV wall thickening and segmental WM applying the AHA model were obtained and statistically analyzed.The mean LV-EDV (r = 0.96) and ESV (r = 0.98) as well as LV-EF (r = 0.97), SV (r = 0.83), and MM (r = 0.95) correlated well. Bland Altman analysis revealed little systematic underestimation of LV-EF (-1.1 +/- 7.8 %), EDV (-0.3 +/- 18.2 ml), SV (-1.3 +/- 16.7 ml) and little overestimation of ESV (1.1 +/- 7.8 ml) and MM (12.8 +/- 14.4 g) determined by DSCT. Systolic reconstruction time points correlated well (DSCT 32.2 +/- 6.7 vs. MRI 35.6 +/- 4.4 % RR-interval). The LV wall thickness obtained by DSCT and MRI showed close correlation in all segments (Ø diff 0.42 +/- 1 mm). In 413 segments (89 %) WM abnormalities were equally rated, whereas DSCT tended to underestimate the degree of wall motion impairment.DSCT with optimized temporal resolution enables regional and global LV function analysis as well as segmental WM analysis in good correlation with MRI. However, the degree of WM impairment is slightly underestimated by DSCT.

German breast cancer screening is monitored by a large physical quality assurance program. This report refers to the first experiences of the Reference Center (RC) Muenster after three years of the technical quality control of digital and analog mammography units (MU). This paper also shows whether the presently used quality assurance (QA) method is able to ensure that the MUs in the screening program are functioning without any serious problems.RC Muenster supervises 95 units (May 2008). The daily, weekly and monthly quality assurance of these units is controlled by web-based QA software named "MammoConrol" and developed by RC Muenster. The annual QA for the units must be conducted in the form of an on-site inspection by medical physics experts of the RC and is scored by an objective ranking system. The results of these QA routines were evaluated and analyzed for this paper.During the period from 3/1/2006 to 5/31/2008, 8 % of the analog systems and 1 % of the digital systems exhibited problems in the daily QA. For 9 % of the analog MUs and 17 % of the digital MUs, failures appeared in the monthly QA. In the annual control, 86.7 % of the analog units exhibited slight problems and 13.3 % had serious problems. With respect to the digital units, 12 % were without any defects, 58 % had slight problems, 27 % had serious failures and 3 % had to be reported to the responsible authorities and were temporarily shut down.The special quality control requirements for German breast cancer screening, including annual on-site checks of the units, have shown in the last three years that QA with a high monitoring standard can be ensured for a large number of decentralized MUs. The currently used QA method sufficiently ensures that the screening program is technically safe. Further studies must show whether the density and focus of the QA measures must be reconfigured.

The spectrum of pathology affecting the central nervous system (CNS) in patients suffering from acquired immunodeficiency syndrome (AIDS) includes not only the human immunodeficiency virus (HIV) infection itself but also opportunistic infections and tumors secondary to AIDS. Despite progress in antiretroviral therapy and the subsequent decrease in the incidence of associated diseases, opportunistic infections and tumors secondary to the HIV infection continue to be the limiting factor in terms of survival with AIDS. Therefore, the therapeutic aim is permanent antiretroviral therapy as well as early diagnosis and treatment of opportunistic infections. Magnetic resonance imaging is often the diagnostic method of choice in suspected CNS pathology of HIV patients. In the following, the typical clinical and radiological features of several AIDS-related pathologies are presented and discussed.

The predicted postoperative forced exspiratory volume in one second (FEV (1)) is an important functional factor for predicting the operability of patients with bronchial carcinoma. A software tool that uses a preoperative chest MSCT and pulmonary function test (PFT) for largely automated prediction of the FEV (1) was evaluated.Fifteen patients with surgically treated lung cancer were examined with a preoperative chest MSCT (1.25 mm slice thickness, 0.8 mm reconstruction increment) and PFT before and after surgery. CT scans were analyzed by the prototype software MeVisPulmo (MeVis gGmbH, Bremen) that predicted the postoperative FEV (1) as a percentage of the preoperative values measured by PFT. The automated segmentation and volumetry of lung lobes were performed either without or with minimal user interaction. Patients underwent lobectomy in twelve cases (6 upper lobes, 1 middle lobe, 5 lower lobes) and pneumectomy in three cases. The predicted FEV (1) values were compared to the observed postoperative values as a standard of reference. The additional functional parameters "total lung capacity" (TLC) and "forced vital capacity" (FVC) were compared to the FEV (1) results.Automated calculation of predicted postoperative lung function was successful in all cases. Due to an implausible PFT, two of the 15 patients were excluded from the collective. A mean postoperative FEV (1) value of 75 % (SD +/- 12 %) of the preoperative FEV (1) was calculated and 74 % (SD +/- 12 %) was actually measured. The deviations of the predicted value from the measured postoperative FEV (1) ranged between - 289 ml (-12 % of the measured postoperative FEV (1)) and + 294 ml (+ 15 % of the postoperative FEV (1)). The mean deviation (absolute value) was 137 +/- 77 ml/s. This corresponds to 7 +/- 3 % of the measured postoperative FEV (1). Bland-Altman-Statistics showed the 95 % "limits of agreement" for the predicted FEV (1) values to be between - 341 ml and + 301 ml, corresponding to - 17.5 % and + 15.8 of the measured postoperative FEV (1) value. Analysis of the TLC and FVC yielded similar results.In the present pilot study the software-assisted prediction of the postoperative FEV (1) using a preoperative MSCT and pulmonary function test corresponded satisfactorily with the observed postoperative values. The introduced approach may make it possible to obtain additional information for the prediction of functional operability prior to performing lung cancer surgery.

To evaluate a new coronary MR angiography technique covering the whole coronary artery tree in one data set acquisition.Six healthy volunteers and 15 patients with known CAD were examined with a navigator gated and corrected (NAV) free-breathing 3D steady-state free precession sequence covering the whole heart (WH-MRA) (TR = 5.4, TE = 2.7, SENSE factor = 2, 160 slices, 0.75 mm reconstructed slice thickness, in-plane resolution = 0.99 x 0.99 mm(2), scan time 14 min [50 % NAV efficiency]) and a vessel targeted 3D SSFP MRA sequence (t-MRA) (TR = 5.6 ms, TE = 2.8 ms, 20 slices of 1.5 mm reconstructed slice thickness, in-plane resolution = 0.99 x 0.99 mm(2), scan time = 7 min [50 % NAV efficiency]). Subjective image quality (4-point scale) and objective image quality parameters including vessel sharpness, vessel diameter and CNR were calculated for WH-MRA and t-MRA. In patients, the accuracy for detection of stenosis larger than 50 % was compared to the accuracy of X-ray coronary angiography (XA), which was considered the standard.WH-MRA demonstrated good vessel visibility in healthy subjects (100 %) whereas vessel visibility in patients was limited (78 % in an 8 segment evaluation). Vessel sharpness was inferior to that of t-MRA in patients (37 vs. 42 %) but equal in healthy subjects (42 %). Vessel diameter did not differ significantly between WH-MRA and t-MRA. CNR was significantly reduced for WH-MRA (CNR 7.4 vs. 11.5). The diagnostic accuracy for the detection of CAD was comparable for both MRA approaches (85.5 vs. 86.2 %).WH-MRA allows good coronary artery visualization in healthy subjects and patients and provides a simplified scanning procedure and advantages in 3D post-processing. Regarding image parameters and the detection of CAD, the results are comparable to those acquired with t-MRA. The major disadvantage remains the high number of diagnostically insufficient images.

Aim of this prospective study was to investigate the incidence of spontaneous cervical artery dissection (sCAD) and cerebral ischemia in patients with suspected sCAD by using a combined head-neck MR-imaging protocol.51 consecutive patients (24 m, 27 f, mean age 39.5 years, range 18 - 55 yrs) admitted to our stroke unit with suspected sCAD according to clinical criteria and age55 years underwent a combined head and neck MR examination within 24 hours of admission (Gyroscan Intera 1.5 T, Philips). Head MRI included ax FLAIR, ax T (1), ax DWI and TOF angiography (imaging time 12 min). Neck MRI consisted of ax T1w-TSE, T2w-TSE, contrast enhanced T1w-TSE and CE-MRA (imaging time 17 min). Three radiologists assessed both studies in consensus with regard to the presence of sCAD and acute ischemia.One patient had to be excluded because of motion artefacts. In 17 of 50 patients sCAD was diagnosed, and in 20 of 50 patients cerebral ischemia. In 5 patients cerebral ischemia was caused by sCAD.The proposed combined MR protocol allows imaging work-up of patients with suspected sCAD within approximately 30 min, resulting in conclusive information about the status of the extracranial vasculature and the presence of ischemia. The high incidence of patients with definite sCAD and the low incidence of cerebral ischemia indicates the necessity of an early definite diagnosis in order to start timely anticoagulation to prevent development of stroke.

To evaluate the applicability of the Canadian CT Head Rule (CCHR) on head trauma patients in a German university hospital.122 patients (m = 74; f = 48; 40 +/- 19 years) were examined with cranial CT due to minor head trauma. The need for cranial CT according to the CCHR was evaluated retrospectively.With a sensitivity of 98.9 % and a specificity of 46.6 % all patients with the need for neurosurgical intervention were detected by applying the major criteria of the CCHR. Also, every patient with severe brain injury was detected by the extended criteria with a sensitivity of 99.6 % and a specificity of 34.1 %. This would have led to a reduction in the rate of cranial CT examinations by 45.1 % for the major and 22.1 % for the extended criteria. No patient with severe brain injury would have been missed by application of the criteria.The Canadian CT Head Rule for patients with minor head trauma is applicable with a very high sensitivity and the potential of significantly reducing the rate of cranial CT examinations in these patients.

The increased T1 relaxation times at 3.0 Tesla lead to a reduced T1 contrast, requiring adaptation of imaging protocols for high magnetic fields. This prospective study assesses the performance of three techniques for T1-weighted imaging (T1w) at 3.0 T with regard to gray-white differentiation and contrast-to-noise-ratio (CNR).Thirty-one patients were examined at a 3.0 T system with axial T1 w inversion recovery (IR), spin-echo (SE) and gradient echo (GE) sequences and after contrast enhancement (CE) with CE-SE and CE-GE sequences. For qualitative analysis, the images were ranked with regard to artifacts, gray-white differentiation, image noise and overall diagnostic quality. For quantitative analysis, the CNR was calculated, and cortex and basal ganglia were compared with the white matter.In the qualitative analysis, IR was judged superior to SE and GE for gray-white differentiation, image noise and overall diagnostic quality, but inferior to the GE sequence with regard to artifacts. CE-GE proved superior to CE-SE in all categories. In the quantitative analysis, CNR of the basal ganglia was highest for IR, followed by GE and SE. For the CNR of the cortex, no significant difference was found between IR (16.9) and GE (15.4) but both were superior to the SE (9.4). The CNR of the cortex was significantly higher for CE-GE compared to CE-SE (12.7 vs. 7.6, p0.001), but the CNR of the basal ganglia was not significantly different.For unenhanced T1 w imaging at 3.0 T, the IR technique is, despite increased artifacts, the method of choice due to its superior gray-white differentiation and best overall image quality. For CE-studies, GE sequences are recommended. For cerebral imaging, SE sequences give unsatisfactory results at 3.0 T.

To evaluate hygienic conditions using automatic injectors in magnetic resonance imaging (MRI) and computed tomography (CT) during clinical routine.The surfaces of medical devices (e. g., control console) and the palms of the technical and medical staff were microbiologically analyzed by taking imprints before and after hygienic education. In addition, the injector syringes for contrast medium (CM) and saline were checked for microbiological contamination following multiple (MRI: 14 h; CT 8 h) and single use. Furthermore, the potential of retrograde contamination from the patient along the tube was analyzed.A bacterial contamination with typical dermal bacteria was documented for the surfaces of the medical devices, the palms of the technical and medical staff, and the injection syringes following multiple use (MRI: 10/10 CM syringes, 6/10 saline syringes; CT: 8/10 CM syringes, 5/10 saline syringes). Correct hand disinfection in combination with single use of syringes avoided bacterial colonization. Retrograde bacterial contamination from the patient was not observed.Regular hygienic teaching sessions for technical and medical staff in MRI and CT departments using automatic injectors should be mandatory. Furthermore, the multiple use of syringes should be avoided until investigations addressing the potential of bacterial contamination are performed.

To evaluate a software algorithm for automated localization of pulmonary nodules at follow-up CT examinations of the chest and to determine factors influencing the rate of correctly matched nodules.The "real-time automatic matching" (RAM) algorithm (Siemens LungCare trade mark software) was applied to 22 follow-up multirow-detector CT (MDCT) examinations in 11 patients (Siemens Somatom VolumeZoom, tube voltage 120 kVp; effective tube current 20 mAs (n = 18) or 100 mAs (n = 4); 4 x 1 mm detector configuration, 1.25 mm slice thickness; 0.8 mm reconstruction increment; standard lung kernel B50f) with a total of 190 lung nodules (mean diameter 6.7 +/- 3.5 mm, range 2 - 17 mm). The following nodule features were recorded: diameter, edge definition (well- or ill-defined), location (upper, middle or lower third; central or peripheral; right or left lung) and inspiration level (considered identical if the difference of diaphragm-apex distance between baseline and follow-up examination was5 %, otherwise it was considered different). A nodule was regarded as correctly localized if the marking box drawn by the software was visible on at least one slice together with the nodule and the center of the nodule was located inside the marking box. chi(2)-test was used to describe influence of nodule features on detection rate. Influence of nodule size was assessed using Mann-Whitney-U-Test.RAM correctly located 164 of 190 of all lung nodules (86.3 %). Detection rate did not depend on nodule location (left vs. right lung: p = 0.48; upper vs. middle vs. lower third: p = 0.96; peripheral vs. central: p = 0.47) or diameter (p = 0.30). Influence of inspiration level was highly significant (p0.001): nodules were detected in 100 % (146/146) for identical inspiration levels and in 40.9 % (18/44) for different inspiration levels. The observation of a significant better localization of ill-defined nodules (p = 0.028) corresponds to a statistical artifact due to the inhomogeneous distributions of this specific feature in our data.RAM is a valuable tool for follow-up of lung nodules at CT. Only very different inspiration levels influenced detection rate.

The so-called non-occlusive disease (NOD) or non-occlusive mesenteric ischemia (NOMI) is a severe and life-threatening pathology. Even under optimal circumstances and standardised diagnostic and therapeutic procedures maximum survival rates do not exceed 50 %. The NOD is a pathology of the elder patient and its incidence rises with other comorbidities such as reduced cardiac output, diabetes and renal insufficiency. Induction of the disease with a severe vasoconstriction of the splanchnic vessels may be a simple cardiac decompensation, a frequent trigger however is a previous heart surgery with consecutive cardiac shock. Early diagnosis is difficult to conduct because of unspecific symptoms. Beside abdominal pain in awake patients, ileus or subileus is remaining the single acute symptom which could be also a consequence of a postoperative paralysis. Laboratory parameters such as leucocytosis and elevated lactat levels are often positive, but unspecific and the latter may be a delayed sign of progressive disease. The only sufficient method for diagnosis implicating a possible treatment option seems to be an immediate angiographic examination. Because of the disappointing results of a solitary surgical approach transarterial medication via catheter is indicated. Depending of the course of the disease only a combination of local mesenteric infusion of vasodilatory drugs and surgical resection of already necrotic bowel promises a successful therapeutic approach and better survival rates.

To analyze prospectively abdominal ultrasound in an emergency department (ED) with special emphasis on different routes of obtaining radiology service, reasons for consultation and ordering diagnosis, and to identify strategies for realizing cost savings.Over a three-month period, the course of events of 295 patients, who entered the ED and underwent on-call sonography, was recorded from the initial presentation to the commencement of treatment, and the results were subsequently evaluated.Of all patients sonographically examined in the ED, 60 % had no prior contact to ambulatory medical services (self-referred, S), 15 % had been sent by a physician (physician-referred, P), and 25 % were inpatients. The most frequent reason for consultation was acute abdominal pain in self-referred patients (70 %) and blunt abdominal trauma in physician-referred patients (70 %). Sonography for acute abdominal pain was considered indicated by the examining radiologist prior to the examination in 90 % (P) and 70 % (S). Sonography contributed to the diagnosis in 44 % (P) and 22 % (S). Patients had to be hospitalized in 80 % (P) and 40 % (S), and needed immediate treatment in 78 % (P) and 43 % (S). Sonography after blunt abdominal trauma was considered appropriate in 90 % (P) and 65 % (S). Diagnostic information was obtained in 90 % (P) and 88 % (S), and therapeutic interventions were required in 10 % (P) and 0 % (S). The admission rate was 97 % (P) and 29 % (S), respectively.In the ED, sonography was performed with significantly lower clinical effectiveness on self-referred than on physician-referred patients. Substantial cost-savings could be achieved by more selective use of abdominal sonography for self-referred patients.

To evaluate a possible dose reduction in pediatric pelvic radiographs in congenital hip dysplasia using a digital flat-panel system instead of a phosphor-storage system.During a six-month period, all pediatric patients referred for pelvic radiography for the evaluation of congenital hip dysplasia were randomly assigned to be examined by either a phosphor-storage system or a digital flat-panel system, whereby the latter system was operated with half the radiation dose. Thirty pairs of radiographs were assessed for the visibility of 16 anatomic details and for 5 orthopedic-radiographic measurements (5-point scale with 1 = excellent; three independent observers). The projection indices of Ball and Kommenda and of Tönnis and Brunken were calculated for all radiographs. The Student's t-test was used to compare the flat-panel and the phosphor-storage radiographs for observers' assessments, patients' age and projection indices.In a total of 7560 observations, the scores for the visibility of anatomic details and orthopedic-radiographic measurements were respectively 2.72 and 2.64 for the flat-panel system and 2.93 and 2.79 for the phosphor-storage system. No significant differences were found between both systems (p0.05) and between patient age and projection indices (p0.05).Pediatric pelvic radiographs can be obtained with a digital flat-panel system using half the radiation dose instead of a phosphor-storage system without sacrificing relevant information in the diagnosis of congenital hip dysplasia.

To compare selenium-based digital radiography with conventional screen-film radiography of the cervical spine.In a prospective study X-ray images of the cervical spine were obtained in 25 patients using selenium- based digital radiography and conventional screen-film radiography. All images were clinically indicated. Selenium-based digital radiography and conventional screen-film radiography were used in a randomized order. Four radiologists independently evaluated all 50 examinations for the visibility of 76 anatomic details according to a five-level confidence scale (1 = not visible, 5 = very good visibility). From the evaluation of these anatomic details scores for the upper and middle cervical spine, the cervicothoracic junction and the cervical soft tissues were calculated. The scores for selenium-based digital radiography and conventional screen-film radiography were compared using Wilcoxon's signed rank test.From a total of 15,200 observations (608 per patient) the following scores were calculated for selenium-based digital radiography and for screen-film radiography, respectively: Upper cervical spine 3.88 and 3.94; middle cervical spine 4.60 and 4.48; cervico-thoracic junction 3.64 and 2.62; cervical soft tissue 4.47 and 3.46. The differences between the last two scores were statistically significant (p0.05).The use of selenium-based digital radiography is superior to conventional screen-film radiography in the depiction of anatomic details of the cervicothoracic junction and the cervical soft tissues.

To prospectively determine the value of magnetic resonance imaging (MRI) with flow quantification in the portal vein for the follow-up of patients with transjugular intrahepatic portosystemic shunt (TIPS).Thirty-six patients with TIPS (23 m, 13 f) were evaluated with MR of the liver parenchyma and quantification of flow in the portal vein. MR examinations were correlated with Doppler sonography and conventional angiography including measurement of the portal pressure gradient (PPG). In cases of re-interventions (dilatation/stent application) additional examinations with MRI and Doppler sonography were performed.MR flow measurements in the portal vein correlated with Doppler sonography (r = 0.69) whereas no correlation of both methods with the PPG was found. No threshold velocity in the portal vein could be determined to predict shunt stenosis. All shunt occlusions (n = 5) were diagnosed correctly by MRA. Thirty measurements before and after successful angiographic interventions revealed a significant increase in portal flow velocity and a significant decrease of the PPG. Magnetic resonance images enabled a reliable detection of procedural complications (parenchymal bleedings, n = 31; extra and subcaspular hematomas, n = 2 each) and newly occurring hepatocellular carcinomas (n = 2) in the follow-up period.Magnetic resonance imaging in the follow-up of TIPS enables a morphological assessment of the liver and an accurate velocity mapping, but is not suited to predict shunt dysfunction as a single method.

This paper reviews the evaluation of the breast of women by MR-techniques after implantation with silicon gel protheses. The main topics are the diagnosis of implant defects such as extensive "gel bleed" and intra- and extracapsular ruptures. Moreover, the MR-detection of siliconomas (encapsulated silicone) and differentiation from malignomas as well as MR-features of chronic foreign body reactions are presented. "Gel bleed" is difficult to diagnose unambiguously by MRI alone. The "linguini" sign is the only reliable mans to diagnose intracapsular ruptures. The presence of silicone outside the implant capsule indicates extracapsular rupture. The MR-spectroscopic detection of silicone in the liver suggests after short implantation times and a normal MR scan the diagnosis "gel bleed", and after longer implantation times of more than 10 years and missing "linguini" sign the diagnosis of ruptures due to a dissolved shell of the implant. MRI, in comparison to other imaging modalities, has the highest specificity and sensitivity in the diagnosis of implant defects. Due to its high costs, however, MR is not suitable as a screening tool and should only be used in cases of sonographic suspected rupture or after radical mastectomy. In these cases MRI is the method of choice.

To evaluate the image quality and radiation exposure of different spiral CT scanning parameters for routine staging examination of the abdomen in oncologic patients using a multi-slice CT scanner.Examination of 40 patients in 4 groups on a multi-slice CT scanner (Somatom VolumeZoom, Siemens AG, Forchheim). Functional detector width (4 x 2.5, 4 x 5 mm) and pitch (table feed in relation to collimated slice width) were varied (3 and 5). Tube voltage (120 kV), effective tube current (160 mAs), slice-thickness (6 mm), increment (4 mm), kernel (B 30), and contrast injection parameters were kept constant. Axial images were assessed by three radiologists regarding delineation of anatomic structures, artifacts, and overall image quality.Significantly reduced image quality especially due to artifacts was observed using a 5 mm detector configuration with a pitch of 5 (scan time 9 sec). Image quality was rated best for a 2.5 mm detector configuration with a pitch of 3 and a scan time of 28 sec. The effective dose was independent of the pitch. However, the mean effective dose was 9% higher using the smaller detector configuration (9.9 mSv vs 10.9 mSv).For routine staging CT of the abdomen use of a 4 x 2.5 mm detector configuration with a pitch between 3 and 5 is recommended. A 4 x 5 mm detector configuration using overlapping data acquisition can also be recommended, but additional thin slice reformations are not possible.

To evaluate left ventricular (LV) myocardial function in ten patients with coronary artery disease (CAD) preoperatively and 6 months after coronary bypass grafting (CABG) by cardiac MRI.Ten patients (mean 65.2 +/- 5.9 years) with angiographically proven CAD and an indication for elective CABG underwent prospective evaluation of global LV function and regional wall motion by Cine-MRI at rest using a multiphase FLASH-2D sequence following regions of interest (ROI)-defined diagnostics of regional myocardial wall motion by means of levocardiography. Within the ROIs a total of 613 LV myocardial segments were analyzed preceding and following surgical revascularization. Results were compared with the data of 10 healthy volunteers.Preoperatively, patients showed reduced stroke volume and ejection fraction compared with volunteers (p0.01). Enddiastolic wall thickness (EDWT) and systolic wall thickening (SWT) were significantly lower in the patients (p0.01). Based on preoperative levocardiography ROI-defined myocardial segments showed a significantly lower preoperative EDWT in areas with wall motion abnormalities (7.4 +/- 2.5 mm; p0.01) than in normal myocardium (9.2 +/- 2.1 mm). Ejection fraction (p0.05), endsystolic wall thickness, and SWT (p0.01) improved significantly after bypass surgery. On ROI-defined analysis myocardial segments with impaired preoperative wall motion (n = 243) showed a significant increase of EDWT, ESWT and SWT (p0.01).In patients with CAD, cardiac MRI enables the non-invasive determination of postinfarctional LV remodeling with an increased EDWT of myocardial segments with normal regional wall motion and of the improvement in global and regional myocardial function following coronary bypass surgery.

To establish recommendations for diagnostic imaging of cystic lesions of the sacral region and to evaluate for potential predictors of therapeutic outcome.Conventional imaging of the lumbar spine, conventional CT, myelography, post myelo-CT and MRI were performed in 7 symptomatic [corrected] patients. All patients underwent operative decompression and histological examination. The radiographic examinations were reevaluated retrospectively and correlated with the postoperative outcome.In one patient the differential diagnosis of a neurinoma could not be excluded with CT and CT-myelography alone. In all 7 patients MRI provided a definite diagnosis and a precise presentation of the cyst extension. A complete postoperative remission of symptoms was noticed in 4, a partial remission in three cases. The correlation of cyst extension, cyst shape, and the communication with the subarachnoid space did not provide predictive information concerning operative outcome.MR can be used as the sole imaging tool for demonstration of cystic lesions. Relevant disadvantages in comparison to myelography and myelo-CT were not evident. No imaging modality could predict the value of surgical intervention.

To define MR imaging characteristics of primary aneurysmal bone cyst.MR imaging studies of 38 patients with histologically proven primary aneurysmal bone cyst were reviewed with reference to morphological features, signal characteristics, and patterns of contrast-enhancement.Most lesions were well marginated towards bone and soft tissues (95%), either surrounded by a complete (84%) or incomplete (16%) rim of low signal intensity on images of all pulse sequences. Frequent features were polycyclic margins (84%), cortical expansion (87%), cystic spaces (100%), contrast-enhancing cyst walls (100%), internal septations (89%), fluid levels (71%) and diverticula-like projections of cyst walls (68%). Solid tissue components could be identified by MR imaging in all lesions which, on pathological examination, contained larger portions of solid material (18%). Edema of surrounding soft-tissues was observed in 29% of the cases.Primary aneurysmal bone cysts demonstrate a relatively uniform MR imaging appearance, which reflects the patho-anatomic composition of the lesion.

The demonstration of calcification of the coronary arterial wall indicates the presence of coronary heart disease (CHD). The prevalence of coronary calcifications increases with age. The extent of the calcifications correlates with the total coronary plaque burden and with the probability of a future myocardial infarction in symptomatic patients. In asymptomatic subjects with risk factors for a myocardial event demonstration of coronary calcifications is diagnostic for coronary atherosclerotic disease before clinical manifestation of the disease. Exact quantification of coronary arterial calcifications (calcium scoring) has become possible with electron beam computed tomography (EBCT) or ECG triggered subsecond CT scanners. Further improvements in the detection of coronary calcifications can be expected with the introduction of multi-slice CT. The prognostic relevance of coronary calcium scoring in asymptomatic high-risk patients is not yet clearly defined. It remains to be clarified whether newer, volume based methods of calcium quantification will be superior to the classic calcium score (Agatston-Score) for risk assessment and follow-up in this patient group.

Purpose of this review is to demonstrate typical X-ray, CT and MR morphology of primary bone tumors and "tumor-like lesions" of the shoulder in correlation with histopathology. 711 primary bone tumors of the shoulder and proximal humerus were studied. 602 were localized in the humerus, 90 in the scapula and 19 in the clavicula. The most frequent benign tumors were osteochondromas (n = 143), simple bone cysts (n = 115), enchondromas (n = 75) and aneurysmal bone cysts (n = 50). Fibrous dysplasia (n = 25), chondroblastoma (n = 15), osteoid osteoma (n = 13), giant cell tumors (n = 12) and non ossifying fibroma (n = 11) were less frequent. The most frequent malignant bone tumors were osteosarcoma (n = 72), chondrosarcoma (n = 52) and Ewing's sarcoma (n = 46). Focal plasmocytoma (n = 20) and lymphoma (n = 10) were less frequent. The average age of the patients was 31.5 years. Some of these tumors were typically located in the shoulder, i.e. simple bone cysts and chondroblastoma. In summary the shoulder was a rather infrequent site of primary bone tumors, but since most of these tumors were benign, the radiologist should be aware of the differential diagnosis to guide therapy.

To investigate the normal sequential MR-pattern of uncomplicated fracture healing within the first 6 weeks.In 8 patients with distal radius fracture, 4 by 4 with and without intraarticular fracture, MR examinations were performed 3-4 days after the onset of fracture and after 1, 2, 3, 4, 5 and 6 weeks using the following techniques: T1-weighted spin echo, T2-weighted spin echo, T2 proton density weighted, STIR (short inversion time inversion recovery), T2-weighted turbo spin-echo and fast field echo (FFE, gradient echo). Sequential examinations were analysed regarding 1) the appearance of fracture line, 2) development of bone marrow signal, and 3) signal changes in surrounding soft tissue. Maximum contrast of bone marrow and soft tissue was measured.The fracture line in bone marrow was best detectable in FFE- and T2-weighted spin ech images during the first week and in T2-weighted fast spin echo and proton density-weighted images from the 2nd to 4th week. Beginning in the 5th-6th week an increase of signal intensity in the fracture gap was typical in all sequences except for T1-weighted spin echo. The contrast maximum of bone marrow and soft tissue, best detectable in STIR- and T2-weighted spin echo images, occurred in the 1st-2nd week.MRI shows a typical time-depending pattern of the fracture line, surrounding bone marrow and soft tissue in normal fracture healing. The degree of injury influences the development of signal changes.

To investigate the effect of direct current treatment (DCT) on the growth of mammary carcinomas in rats by MR-volumetry.Chemically induced mammary adenocarcinomas in a control group (n = 17) were compared with treated tumours (18 C/cm3 in group A: n = 7 or 36 C/cm3 in group B: n = 12). 31 untreated tumours were situated near a treated tumour (group C). Experiments were carried out using one positive electrode in the tumour centre and three negative electrodes in the periphery. The tumour volume was measured by MRI before, and 1, 3, 6, 9 and 12 weeks after treatment.12 weeks after DCT, the mean tumour volume in group A (164% +/- 158%, p0.05) and group B (13% +/- 24%, p0.001) was significantly reduced compared to the control group (434% +/- 230, Mann-Whitney U-Test). Complete tumour regression occurred in 42% of tumours in group B and was not achieved in group A, C and control group. Tumour growth in group C was decreased compared to the control group.The effectiveness of DCT was found to depend on the applied dosage -36 C/cm3 was more effective than 18 C/cm3. The effect of DCT is not limited to the area between the electrodes.

To compare the accuracy of high resolution endorectal magnetic resonance imaging (EMRI) and endorectal ultrasound (EUS) in the preoperative diagnostic of rectal tumours.Twenty-one patients with known rectal tumours underwent MR imaging with an endorectal surface coil and EUS. Transversal EMR images were obtained using fast T2-weighted sequences and pre- and postcontrast T1-weighted images. EUS was performed using a 7.0 MHz transducer. Results of both methods were compared with specimens from the resected tumours.Rectal wall layers were reliably demonstrated with both methods in all patients. EMRI and EUS determined the depth of rectal wall invasion. EMRI and EUS agreed with pathologic findings in 16/21 cases, respectively. In one case each, both methods understaged one tumour. EMRI overstaged an adenoma as a T2-tumour. In three and four patients, respectively, no staging was possible due to technical problems.EMRI and EUS show comparable results in the preoperative T-staging of rectal tumours. Both techniques are not suitable for differentiating benign from malignant lymph nodes accurately. While EMRI is expensive and technically demanding, it allows an objective documentation of pathological findings which is less dependent on the examiner. Two important clinical conclusions can be drawn from the good results of T-staging: Adenomas and T1-tumours can be treated by local excision. In patients with advanced tumours (T3/T4) a neoadjuvant therapy can be initiated.

To evaluate accuracy of needle localisation using and signal enhancement on a 1.OT MR imager for various needles for MR-guided biopsy.The differences between actual and virtual needle position of needles with different orientations were evaluated in a phantom for spin-echo including turbo-spin-echo sequences.Artifacts depended on the orientation of the needle relative to the field B0, frequency-encoding gradients (Gf) and slice orientation. This resulted in different artifact shapes and sizes for left or right and cranial or caudal biopsy access routes. Applying turbo spin echo sequences feasible for biopsy, the signal void of a 18 G needle (Cook) parallel to Gf reached between 0.3 and. 4.6 mm further into the medium than the real needle tip, depending on needle orientation relative to B0. The diameter of the signal void around the needle varied, the needle shaft was right in the centre of the signal void. With Gf orthogonal to the needle the offset of signal void to needle tip ranged from 2.7 to 3.3 mm, while the actual position of the needle shaft was up to 3.3 mm lateral of the signal void center. While nominal echo times did not influence the size of the artifact in turbo-spin-echo sequences, the artifacts increased with smaller matrix and larger water-fat shift. Material and mass of the needle determined the size of the artifacts as well.Localisation accuracy of the needle can be optimised by choosing optimal gradient directions depending on whether needle tip or shaft position should be displayed.

To evaluate whether phosphorus magnetic resonance spectroscopy (31P-MRS) enables a non-invasive detection of liver involvement in systemic diseases like Hodgkin's lymphoma.Using a clinical 1.5 Tesla whole-body MR system image-guided localised phosphorus MR spectra from the anatomically defined volumes of interests were measured. A combination of surface coil, adiabatic excitation pulse and modified image-selected in vivo spectroscopy (ISIS)-sequence was applied. The spectroscopy data were evaluated quantitatively with a time-domain fit programme using non-linear optimisation algorithms to quantify peak areas. After establishment of the examination protocol, 22 healthy volunteers and 13 patients with suspected lymphoma infiltration of the liver were examined.Liver spectra of patients suffering from lymphoma infiltration differed significantly from spectra of persons with normal liver: 1. The peak area ratio of phosphomonoesters (PME) to beta-NTP was elevated in all patients with histologically confirmed liver lymphoma. 2. Patients suffering from Hodgkin's disease with specific or unspecific liver infiltration (n = 7) could be differentiated from patients without liver involvement. In case of infiltrated liver, the peak area ratio PME to beta-NTP was increased, and the pH value was shifted to lower values. Unambiguous differentiation between non-specific (n = 3) and specific (n = 4) infiltration of the liver was not possible. 3. In patients after cytostatic treatment (n = 3), an increase of the peak area ratio of inorganic phosphate to beta-NTP was observed.Our preliminary results indicate that 31P-MRS can yield pointers to liver involvement in patients with systemic diseases such as Hodgkin's disease, which may be hardly detected by imaging methods.

Analysis of the MRT signals and their extent from intra- and paraspinal abscesses with reference to predisposing factors, their causes and localisation.The histories and MRT findings in 34 Patients with intra- and paraspinal abscesses were evaluated retrospectively. Most of the patients (24/34) were older than 50 years. A second peak was below 30 years.27/34 patients had some underlying disease which predisposed to infection, e.g., diabetes mellitus. The most common causal organisms were Staph. aureus (53%) and streptococci (15%). In 23/34 cases (68%), the abscesses were in the thoraco-lumbar or lumbar region, while only 6/34 occurred in the upper two-thirds of the thoracic spine and only 5/34 in the cervical region. In only 8/34 was the abscess confined to two vertebral lengths; in the remaining patients it was much more extensive. Intraspinal abscesses were about twice as large as the vertebral components and 1.5 times greater than paravertebral abscesses. The age of the abscesses could be estimated approximately from the signals.Contrast enhanced MRT permits detailed analysis of the compartments and exact estimation of the extent of the lesions and permits accurate monitoring of treatment.

To determine which appearances on CT or MRI are indications for surgical treatment of intracranial arachnoid cysts.The images obtained from 26 patients treated either by surgery or conservatively were compared retrospectively. The features evaluated were the size of the cyst, evidence of a space-occupying lesion, cerebral abnormalities near the cyst and skull deformities.In 6 patients (23%), surgical treatment was indicated because of the presence of obstructive hydrocephalus. The presence of all other radiological findings did not differ significantly between the two analysed groups. 92% of the patients treated surgically benefited from the treatment. Fenestration and shunting produced similar results in respect to reduction in size of the cyst and the clinical and neurological symptoms.The only indication for surgery is the presence of obstructive hydrocephalus. The other features were seen equally amongst the patients treated conservatively and the surgical patients and had no influence on the decision to operate.

What is the percentage of non-indicated examinations in the routine work of a radiologic university clinic?From April 1st, 1994, to June 30th, 1994, all examinations carried out in the Radiology Department of the University Hospital of Cologne, Germany, were evaluated prospectively by the radiological medical staff using a questionnaire. The question was whether the examinations were medically indicated or not. Requests for examinations which were not carried out due to lacking indications were excluded from evaluation. The percentage of error was assessed by analysis of random samples assigned to the referring clinicians.1400 (5.4%) of a total of 25718 examinations were considered by the radiological medical staff as "not indicated". A subsample analysis performed for validation gave an upper margin of a 95% confidence interval of about 13% for this rating. The percentage of non-indicated examinations was higher in conventional x-ray examinations and ultrasonography than in x-ray angiography, computed tomography and magnetic resonance tomography. Examinations during night time and weekend showed a higher percentage of lacking medical indications than examinations during daytime.Considering a university clinic concerned with educational work, the percentage of radiological examinations carried out without medical indication was tolerable. However, indications must be further improved by intensifying the communication with the referring clinicians and by elaborating a higher standard of radiological and clinical postgraduate training.

Radiological findings and course of progressive multifocal leukoencephalopathy in 14 patients (1 woman, 13 men; 13 HIV seropositive, 1 chronic lymphatic leukaemia) were analysed retrospectively and correlated with clinical symptoms. A total of 21 CT and 16 MRI studies were evaluated. CT scans and MR images of 9 patients, which had been obtained in less than two weeks, could be compared to each other. MRI was superior to CT: 6 lesions with a diameter of 1 cm and below were not detected on CT scans, in 5 patients the extent of lesions was underestimated. Cortical involvement, mass effect or signs of atrophy were missing. Only 1 of 67 lesions showed a tiny enhancement after Gd injection. Due to the pattern and spread of lesions, which showed a close correlation to the neurologic symptoms, three different types of PML are suggested: 1. initial precentral demyelinisation with contralateral hemiparesis (n = 8); 2. lesions in temporo-occipital locations with visual disturbances (n = 2); 3. predominantly bilateral lesions of cerebellar white matter with ataxia (n = 4).

One hundred and forty-six intracranial hematomas in 129 patients were examined by MRI (136 examinations) and CT (147 examinations). Even using a high field MR system (1.5 T) and gradient-echo sequences, CT was the more sensitive method during the acute phase (46% compared with 93%). During the sub-acute phase, MR was superior to CT (97% compared with 58%), as it was in the chronic phase (93% compared with 17%). Petechial bleedings and discreet foci of contusion could only be demonstrated by MRI. Moreover, MRI showed evidence of residues from hemorrhage (signal reduction due to hemosiderin deposition) long after CT has become normal. Analyses of MRI images allows one to date the bleed and to distinguish between a) the formation of clot in a haemorrhagic cavity, b) bleeding into the tissues and c) a liquefying hematoma. Bearing in mind this classification, the localisation of the hematoma and the clinical findings mostly allow it is possible to determine the cause of the bleeding. Another advantage is the certain detection of vessels supplying arterio-venous malformations and cavernous hemangiomas.

A new method for image-guided localized phosphorus NMR spectroscopy of superficial tissues has been investigated using a 1.5 Tesla whole-body-MR-System. We used a surface coil combined with adiabatic excitation pulses and a modified ISIS sequence. This approach is related to imaging sequences and thus permits a flexible and accurate determination of the volume of interest from "conventional" proton images. The scope and advantages of the method are demonstrated by phantom studies. Clinical applications to the liver, renal transplants, and the mediastinum are described.

Using a 1.5 Tesla whole body system, image-guided localized 31P NMR spectra of normal brain tissue and of various intracranial tumours have been obtained. The spectra of normal brain tissue were reproducible, although further studies are necessary to determine physiological variability. Compared with the normal spectrum, the spectra of tumours usually showed reduced intensity of the phosphocreatine peaks and increased intensity of the peaks of inorganic phosphate, of phosphodiesters and particularly phosphomonoesters. To our limited experience there has been no correlation between the spectra and tumour histology. In follow-up studies of brain tumours undergoing treatment by chemotherapy or radiotherapy 31P MR spectroscopy demonstrated metabolic changes at a time when 1H MR imaging did not show any changes in tumour size or structure. Further investigations have to be done in order to evaluate the usefulness of 31P MRS for detecting therapeutic response in follow-up studies.

In thirty-seven MR examinations of intracranial tumours equivalent sections were obtained in a multi-echo technique before and after intravenous injection of 0.1 mmol. gadolinium DTPA/kg body weight. From this comparison the following preliminary conclusions have been drawn concerning the demonstration of the tumour, its delineation and type: Contrast administration does not unequivocally improve the sensitivity. In 55% of the cases, differentiation between tumour and oedema respectively normal brain tissue was easier after Gd-DTPA. Diffusely infiltrating gliomas remain a problem, since their extent is uncertain with or without contrast medium. The structure of the tumour can already be adequately characterized by the multi-echo technique. In order to diagnose the type of tumour, the criteria which apply to Gd-DTPA are similar to those used for iodine-containing contrast media in CT.

Amongst 1.345 MR examinations of the skull 49 patients were suspected of a cerebral infection. With a knowledge of the clinical situation, the abnormal findings were classified according to their localization, number, extent and distribution. The final diagnoses included meningitis, meningoencephalitis, and abscesses of varying etiology. CT was carried out in 29 patients; it was found that MRI was more sensitive and able to show the lesion at an earlier stage (sensitivity 90% compared with 66%). Observations of proton density and relaxation time combined with morphological criteria and clinical history reduced differential diagnoses. On the other hand, the changes due to intracranial infections lead to an inflammatory reaction that may be similar to the findings in degenerative or even tumorous cerebral lesions. This explains why twelve patients with abnormal MRI findings were erroneously diagnosed as having cerebral infections.