Materials and Methods: Institutional ethics committee approval and written informed consent were obtained. Twenty-seven patients (16 men, 11 women; mean age, 64.4 years ± 14.8 [standard deviation]; range, 26–87 years) suspected of having or known to have peripheral arterial disease underwent first-pass and steady-state MR angiography and DSA. First-pass and steady-state MR angiography were performed in the same patient in the same session and with the same dose of blood pool contrast agent. The most severe stenosis grade of each evaluated segment was measured; sensitivity, specificity, and positive and negative predictive values were calculated at first-pass and steady-state MR angiography, with DSA as the reference standard. The coefficient was used to measure the agreement between first-pass MR angiography, steady-state MR angiography, and DSA.
Results: A total of 334 arterial segments were available for intraindividual comparison of first-pass MR angiography, steady-state MR angiography, and DSA in 27 patients. In 20 (74%) of 27 patients, the stenosis grade of at least one of the evaluated vessels differed at steady-state MR angiography from that at first-pass MR angiography. In total, stenosis grade was judged as higher at first-pass MR angiography than at DSA (overestimation) in 28 of 334 segments and as lower (underestimation) in 15 of 334 segments. The stenosis grade as judged at steady-state MR angiography matched with that at DSA in 334 of 334 vessel segments.
Conclusion: High-spatial-resolution steady-state MR angiography allowed for better agreement with DSA regarding stenosis grade in patients with arterial disease compared with standard-resolution arterial-phase first-pass MR angiography.
Supplemental material:
© RSNA, 2008
]]>Materials and Methods: Ethical permission and written informed patient consent were obtained. The study was HIPAA compliant. Forty-seven patients were included and underwent 1.5-T magnetic resonance (MR) imaging of the knee that was treated surgically. Signal intensity characteristics of the ACL graft were evaluated on sagittal intermediate-weighted and sagittal and axial T2-weighted fast spin-echo MR images. The amount of signal intensity change, femoral and tibial graft tunnel position, and orientation of ACL graft in the coronal plane were assessed. Objective index of graft stability or laxity was performed with arthrometric testing, and subjective function was assessed by using International Knee Documentation Committee (IKDC) scoring.
Results: Increased intrasubstance graft signal intensity was found in 70 % (33 of 47) and in 64% (30 of 47) of patients on intermediate-weighted MR images and T2-weighted MR images, respectively. When present, intrasubstance graft signal intensity changes involved less than 25% of the maximal cross-sectional area of the graft in 70% (23 of 33) of cases on intermediate-weighted acquisitions and in 70% (21 of 30) of cases on T2-weighted acquisitions. No significant association was seen between graft signal intensity changes on intermediate-weighted and T2-weighted images and IKDC score (P = .667 and .698, respectively), arthrometric testing (P = .045–.99), and time since surgery (P = .592 and .610, respectively).
Conclusion: Small amounts of increased intrasubstance graft signal intensity on intermediate- and T2-weighted images can be seen after ACL reconstruction at long-term follow-up of 4 years or longer and do not necessarily correlate with findings of joint instability or functional limitations in patients after ACL repair.
© RSNA, 2008
]]>Supplemental material:
© RSNA, 2008
]]>Materials and Methods: This retrospective study was performed after institutional review board approval and informed consent were obtained. Sixty-three consecutive patients treated for NPC underwent follow-up with both MR imaging and FDG PET/CT. Findings were evaluated according to the TNM classification. Final diagnosis was confirmed at biopsy or imaging follow-up for at least 6 months. Proportions and their 95% confidence intervals were computed; for comparison of data obtained separately from MR imaging and FDG PET/CT and those obtained from their combined use, the McNemar test was used. P < .05 was considered to indicate a statistically significant difference.
Results: There was a trend toward greater overall accuracy of MR over PET/CT in detecting residual and/or recurrent NPC at the primary site; 92.1% (58 of 63 patients) for MR versus 85.7% (54 of 63) for FDG PET/CT (P = .16). Overall accuracy for tumor restaging was 74.6% (47 of 63) for MR and 73.0% (46 of 63) for FDG PET/CT (either modality used alone), but the overall combined accuracy was 92.1% (58 of 63) (all P values < .01).
Conclusion: MR imaging demonstrated a trend toward higher accuracy than did FDG PET/CT in detecting residual and/or recurrent NPC at the primary tumor site. The combined use of MR and FDG PET/CT was more accurate for tumor restaging than when either modality was used independently.
© RSNA, 2008
]]>Materials and Methods: This study was approved by the medical ethics committees of the authors' institutions, and informed consent was obtained from participants' parents or guardians. Forty asymptomatic volunteers underwent MR imaging to detect the appendix. Mean age (± standard deviation) for the 18 boys was 12.8 years ± 2.4 (range, 9–17 years) and for the 22 girls was 12.7 years ± 2.7 (range, 8–17 years). The MR imaging protocol consisted of axial T1-weighted fast spin-echo (SE), axial and coronal T2-weighted fast SE, and axial T2-weighted fat-saturated fast SE sequences. Two observers interpreted results independently, and coefficients were calculated to evaluate agreement. The rate of detection was defined as the percentage of times that the appendix was visualized with at least one sequence. Body mass index was recorded for all participants.
Results: The normal appendix was detected by one observer in 19 cases (48%) and by the other in 20 cases (50%). The rate of appendix detection after observers reached a consensus was 48% (95% confidence interval: 32%, 64%). Interobserver agreement for appendix detection was good for most sequences analyzed. The greatest rate of normal appendix detection was found with the axial T2-weighted fast SE sequence (48%).
Conclusion: The rate of normal appendix detection in children and adolescents at unenhanced MR imaging was 48%, which is similar to computed tomographic but lower than ultrasonographic detection rates.
© RSNA, 2008
]]>Materials and Methods: This HIPAA-compliant retrospective study was approved by the institutional review board. Informed consent was waived. Of 146 consecutive pregnant patients suspected of having appendicitis (mean age, 29 years) who underwent MR imaging, 143 had MR images in which the appendix and cecum were identifiable in the sagittal plane. Two observers reviewed the MR images; findings were agreed upon by consensus. With use of sagittal single-shot fast spin-echo MR images, the cecal tilt angle was calculated as the angle between the imaging table and a line drawn between the cecal tip and the luminal center of the most proximal inflection point in the ascending colon. The location of the appendiceal base relative to the lumbosacral spine was recorded. Statistical analyses were performed by using Spearman and Pearson correlation coefficients to evaluate the relationship among gestational age, appendiceal base location, and cecal tilt angle. Receiver operating characteristic curve analysis was performed to assess the ability of the cecal tilt angle to help differentiate between a high and low appendiceal base level.
Results: Cecal tilt angles showed moderate correlation with appendiceal base levels (Spearman correlation coefficient, 0.44; P < .001) and poor correlation with gestational age (Pearson correlation coefficient, 0.25; P = .002). Regardless of gestational age, cecal tilt angles of at least 90° were predictive of a high appendiceal base level with a specificity of 98% (95% confidence interval: 92%, 100%).
Conclusion: The cecal tilt angle is useful for localizing the appendix in pregnant patients at MR imaging and helps predict the location of the appendix within the right upper quadrant of the abdomen with high specificity, irrespective of gestational age.
© RSNA, 2008
]]>© RSNA, 2008
]]>Materials and Methods: Four atherosclerotic specimens were examined with 64-section dual-energy multidetector CT by using a novel dual-detector "double-decker" design, with stacked high- and low-energy detector arrays with 32 x 0.625-mm collimation, at 140 kVp and 400 mAs, acquiring simultaneous and isopedic low- and high-energy data sets. Additionally, combined-energy data sets were calculated, and an enhancement algorithm was proposed. Cardiac motion was simulated by an anthropomorphically moving phantom, and OCT was used as a reference standard for plaque quantification. Univariate general linear model (GLM) analysis was used to compare sizes of plaque calcifications determined with OCT with those determined with dual-energy multidetector CT, and the significance of factors such as cardiac motion was assessed.
Results: GLM analysis revealed that plaque quantification based on low-, high-, and combined-energy data sets differed significantly from that based on OCT (P < .001). Greater data variation occurred in smaller (<8 mm2) and larger (>12 mm2) calcifications. Comparison of calcified plaque sizes determined with OCT with those determined with the dual-energy multidetector CT enhancement algorithm revealed no significant difference (P = .550). Cardiac activity led to a slight increase in data variation in regard to OCT for corresponding static (mean, 10.2% ± 3.2 [standard deviation]) and dynamic (13.8% ± 4.9) dual-energy multidetector CT data sets.
Conclusion: Dual-energy multidetector CT with novel postprocessing techniques enhanced accuracy of calcified plaque quantification by reducing effects of tissue blooming and beam hardening beyond single-energy multidetector CT.
Supplemental material:
© RSNA, 2008
]]>Materials and Methods: Animal protocols were approved by the Institutional Administrative Panel on Laboratory Animal Care. Lipid-shell perfluorocarbon-filled MBs, targeted to VEGFR2 via anti-VEGFR2 antibodies, were radiolabeled by conjugating the radiofluorination agent N-succinimidyl-4-[18F]fluorobenzoate (SFB) to the anti-VEGFR2 antibodies. These MBs were then injected intravenously into nude mice (n = 4) bearing angiosarcomas, and the whole-body biodistribution of these probes was assessed for 60 minutes by using dynamic micro-PET. Results were compared with ex vivo gamma counting (n = 6) and immunofluorescence staining (n = 6). Control studies in angiosarcoma-bearing mice were performed with injection of the radiolabeled antibodies alone (n = 3) or free SFB (n = 3). A mixed-effects regression of MB accumulation on fixed effects of time and tissue type (tumor or muscle) and random effect of animal was performed.
Results: VEGFR2-targeted MBs rapidly cleared from the blood circulation (50% blood clearance after approximately 3.5 minutes) and accumulated in the liver (mean, 33.4% injected dose [ID]/g ± 13.7 [standard deviation] at 60 minutes) and spleen (mean, 9.3% ID/g ± 6.5 at 60 minutes) on the basis of micro-PET imaging. These findings were confirmed with ex vivo gamma counting. Uptake of targeted MBs was significantly higher (P < .0001) in tumor than in adjacent skeletal muscle tissue. Immunofluorescence staining demonstrated accumulation of the targeted MBs within hepatic Kupffer cells and splenic macrophages. Biodistribution of the radiolabeled antibodies and free SFB differed from the distribution of the targeted MBs.
Conclusion: Dynamic micro-PET allows assessment of in vivo biodistribution of VEGFR2-targeted MBs.
© RSNA, 2008
]]>Materials and Methods: This single-center retrospective analysis of 71 symptomatic fibroids in 66 women was approved by the institutional review board and was HIPAA-compliant. Patients were treated with MR imaging–guided focused ultrasound surgery. The volume of treated fibroid and nonperfused volume (NPV) were calculated with software, while symptom outcome was assessed with a symptom severity score (SSS). Fibroids were classified as hyperintense or hypointense relative to skeletal muscle on pretreatment T2-weighted MR images.
Results: Baseline volume of treated fibroids was 255.5 cm3 ± 201.7 (standard deviation), and baseline SSS was 61.5 ± 14.9. Both pretreatment fibroid signal intensity (SI) and posttreatment NPV predicted 12-month volume reduction independently: Fibroids with an NPV of at least 20% or with low SI both showed significantly larger volume reduction (17.0% ± 13.0 and 17.2% ± 20.1, respectively) than fibroids with an NPV less than 20% or with high SI (10.7% ± 18.2 and no significant change, respectively). Patients whose fibroids demonstrated an NPV of at least 20% also experienced a larger decrease in SSS than did patients with fibroids with an NPV less than 20% (50.1% ± 19.8 vs 32.6% ± 29.9).
Conclusion: Fibroids with low SI on pretreatment T2-weighted MR images were more likely to shrink than were ones with high SI. The larger the NPV immediately after treatment, the greater the volume reduction and symptom relief were. These findings may help both in selecting appropriate patients for MR-guided focused ultrasound surgery and in predicting patient outcome.
© RSNA, 2008
]]>Materials and Methods: This study was approved by the institutional review board. Records were reviewed from 101 consecutive patients who had experienced a recent stroke (onset within the previous 1 month) from a suspected cardioembolic source and who had undergone both 64-section cardiac CT angiography and TEE within 1 week. The numbers of thrombi in the LAA detected with cardiac CT angiography and with TEE were recorded, and the agreement between thrombus detection with CT and with TEE was assessed by using statistics.
Results: Eight thrombi in the LAA were detected with TEE, and 12 were detected with cardiac CT angiography. With TEE used as the reference standard, the overall sensitivity, specificity, and accuracy of 64-section cardiac CT angiography for detecting thrombi were 100% (95% confidence interval [CI]): 63%, 100%), 95% (95% CI: 90%, 99%), and 96% (95% CI: 92%, 100%), respectively. The concordance between LAA thrombus detection with 64-section cardiac CT angiography and with TEE was high: 89 patients with no thrombus at CT or TEE; eight patients with thrombus at both CT and TEE; and four patients with thrombus at CT but not at TEE (overall = 0.779 [95% CI: 0.571, 0.987]).
Conclusion: Sixty-four–section cardiac CT angiography is a noninvasive and sensitive modality for detecting thrombi in the LAA of stroke patients. Although TEE is currently considered the reference standard modality for detecting LAA thrombi, 64-section cardiac CT angiography has the potential to become a useful modality for detection of intracardiac thrombus.
© RSNA, 2008
]]>Materials and Methods: Informed consent was obtained in 11 patients (seven women, four men; average age, 58.6 years) with pain related to non–weight-bearing bone metastases who were subsequently treated with MR imaging–guided focused ultrasound in this research and ethics board–approved study. Efficacy was evaluated by changes in visual analog scale (VAS) scores, in pain medication usage, and in quality of life. Safety of the device was evaluated by recording incidence and severity of treatment-related adverse events up to 3 months after treatment at physical examination and follow-up imaging. Follow-up imaging included contrast material–enhanced MR imaging and unenhanced computed tomography (CT) 1 month after treatment and contrast-enhanced MR imaging 3 months after treatment. Imaging studies were assessed for changes in tumor imaging characteristics and any adverse events associated with MR imaging–guided focused ultrasound treatment.
Results: Twelve lesions were treated in 11 patients. All patients reported progressive decrease in pain in treated regions and reduction in pain medication usage during the 3-month follow-up period. VAS scores averaged 6.0 before treatment and decreased to 0.5 at 3 months (decrease in pain scores, 92%; P < .01). No adverse events were recorded at physical examination or follow-up imaging. The majority of patients with osteolytic metastases had varying degrees of necrosis of the enhancing medullary component of the metastasis at follow-up enhanced MR imaging. Five patients had increased bone density at the site of treated osteolytic metastases at follow-up unenhanced CT at 3 months after MR imaging–guided focused ultrasound.
Conclusion: MR imaging–guided focused ultrasound is a noninvasive technique that allows palliative treatment of bone metastases with little or no morbidity.
© RSNA, 2008
]]>Materials and Methods: Young (three women, three men; age range, 25–35 years) and elderly (four women, two men; age range, 68–72 years) groups underwent MR imaging and proton MR spectroscopic imaging at 3 T in this HIPAA-compliant prospective study and gave institutional review board–approved written consent. Volume of interest was centered on and tilted parallel to hippocampal anteroposterior plane. Absolute N-acetylaspartate (NAA), choline, and creatine levels were obtained in each voxel, with phantom replacement.
Results: Mean NAA, creatine, and choline concentrations in the young group were higher in posterior hippocampus (12.9 mmol/L ± 2.0 [standard deviation], 7.8 mmol/L ± 1.2, 2.3 mmol/L ± 0.4, respectively) than anterior hippocampus (8.0 mmol/L ± 1.1, 6.0 mmol/L ± 1.4, 1.5 mmol/L ± 0.2; P = .005, .02, and .0002, respectively). In the elderly group, mean concentrations were higher in posterior hippocampus (8.6 mmol/L ± 0.9, 5.6 mmol/L ± 0.6, 1.5 mmol/L ± 0.2, respectively) than anterior hippocampus (7.2 mmol/L ± 1.0, 2.4 mmol/L ± 0.3, 1.0 mmol/L ± 0.2; P = .006, .0001, .04, respectively). Mean concentrations were significantly higher in the young group (13.2 mmol/L ± 1.0, 7.4 mmol/L ± 0.8, 2.1 mmol/L ± 0.3, respectively) than in the elderly group (9.0 mmol/L ± 1.0, 5.8 mmol/L ± 0.8, 1.8 mmol/L ± 0.3; P = .0001, .01, .05, respectively). Posteroanterior metabolic gradients differed: NAA decreased faster in the young group (–1.0 mmol/L · cm–1) than the elderly group (–0.7 mmol/L · cm–1); creatine and choline concentrations decreased faster in the elderly group (–0.8 and –0.058 mmol/L · cm–1, respectively) than the young group (–0.16 and –0.008 mmol/L · cm–1, respectively). No left-right metabolic differences were found.
Conclusion: Significant metabolic heterogeneity was observed between groups and along anteroposterior axis of healthy hippocampus in both groups. Age matching and consistent voxel placement are important for correct comparisons of both absolute metabolic levels and metabolite ratios in longitudinal intra- and intersubject cross-sectional studies.
© RSNA, 2008
]]>Supplemental material:
© RSNA, 2008
]]>© RSNA, 2008
]]>Materials and Methods: Institutional review board approval and informed consent were waived for this retrospective study. Forty-four consecutive patients with FAI of the cam type were included (mean age, 30.7 years; range, 16–49 years), including 30 men (mean age, 30.5 years; range, 16–49 years) and 14 women (mean age, 31.4 years; range, 18–48 years). The inclusion criteria were no previous surgery, surgery within 3 months after MR imaging, and availability of a detailed surgical report with acetabular cartilage findings. MR arthrographic findings were assessed independently by two blinded readers. Findings at surgery served as reference standard. Sensitivity, specificity, accuracy, and statistics for interobserver agreement were calculated.
Results: At surgery, acetabular cartilage delamination was seen in 23 (52%) of 44 patients (mean size of cartilage flap from acetabular rim, 7.6 mm; range, 2–30 mm). At MR, patients with fluid signal intensity under the cartilage delamination had a respective sensitivity, specificity, and accuracy of 22%, 95%, and 57% for reader 1 and 30%, 95%, and 61% for reader 2. A hypointense line in the acetabular cartilage on sagittal three-dimensional double-echo steady-state images with water excitation demonstrated moderate diagnostic performance (respective sensitivity, specificity, and accuracy were 70%, 57%, and 64% for reader 1 and 70%, 62%, and 66% for reader 2). Hypointense areas in the acetabular cartilage were quite specific on both coronal intermediate-weighted fat-saturated images (respective sensitivity, specificity, and accuracy were 52%, 90%, and 70% for reader 1 and 74%, 90%, and 82% for reader 2) and coronal T1-weighted images (respective sensitivity, specificity, and accuracy were 35%, 90%, and 61% for reader 1 and 61%, 95%, and 77% for reader 2).
Conclusion: Cartilage delamination is common in patients undergoing surgery for FAI. Fluid under the cartilage delamination is a specific but rare finding. Hypointense areas in the acetabular cartilage seen on intermediate-weighted fat-saturated or T1-weighted images appear to be helpful diagnostic criteria.
© RSNA, 2008
]]>Materials and Methods: This study was conducted under an institutional review board–approved, HIPAA-compliant protocol; the need for informed consent was waived. Nine experienced radiologists rated an enriched set of mammograms that they had personally read in the clinic (the "reader-specific" set) mixed with an enriched "common" set of mammograms that none of the participants had previously read in the clinic by using a screening Breast Imaging Reporting and Data System (BI-RADS) rating scale. The original clinical recommendations to recall the women for a diagnostic work-up, for both reader-specific and common sets, were compared with their recommendations during the retrospective experiment. The results are presented in terms of reader-specific and group-averaged sensitivity and specificity levels and the dispersion (spread) of reader-specific performance estimates.
Results: On average, the radiologists' performance was significantly better in the clinic than in the laboratory (P = .035). Interreader dispersion of the computed performance levels was significantly lower during the clinical interpretations (P < .01).
Conclusion: Retrospective laboratory experiments may not represent either expected performance levels or interreader variability during clinical interpretations of the same set of mammograms in the clinical environment well.
© RSNA, 2008
]]>Materials and Methods: MEDLINE, EMBASE, CINAHL, and Cochrane databases were searched from January 1966 to February 2006. Prospective trials were selected if they (a) compared graded compression US and CT in the same patient population; (b) included more than 10 patients, otherwise, the study was considered a case report; (c) evaluated mainly adults or adolescents; (d) used surgery and/or clinical follow-up as reference standard; and (e) reported data to calculate 2 x 2 contingency tables for graded compression US and CT. Estimates of sensitivity, specificity, and positive and negative likelihood ratios (LRs) for US and CT were calculated. Posttest probabilities after CT and US were calculated for various clinically relevant prevalences.
Results: Six studies were included, evaluating 671 patients (mean age range, 26–38 years); prevalence of acute appendicitis was 50% (range, 13%–77%). Positive LR was 9.29 (95% confidence interval [CI]: 6.9, 12.6) for CT and 4.50 (95% CI: 3.0, 6.7; P = .011) for US, yielding posttest probabilities for positive tests of 90% and 82%, respectively. Negative LR was 0.10 (95% CI: 0.06, 0.17) for CT and 0.27 (95% CI: 0.17, 0.43) for US (P = .013), resulting in posttest probabilities of 9% and 21%, respectively. Posttest probabilities for positive tests were markedly decreased at lower prevalences.
Conclusion: In head-to-head comparison studies of diagnostic imaging, CT had a better test performance than did graded compression US in diagnosing appendicitis. Ignoring the relationship between prevalence (pretest probability) and diagnostic value may lead to an inaccurate estimation of diagnostic performance.
Supplemental material:
© RSNA, 2008
]]>Materials and Methods: The institutional animal care and use committee approved this study. Pigs underwent coronary artery occlusion and reperfusion and served as either control (n = 8) or VM202-treated (n = 8) animals. VM202 was transferred intramyocardially into four infarcted and four periinfarcted sites. Cardiac magnetic resonance (MR) imaging (cine, perfusion, delayed enhancement) was performed in acute (3 days) and chronic (50 days ± 3 [standard error of the mean]) infarction. Histopathologic findings were used to characterize and quantify neovascularization. The t test was utilized to compare treated and control groups and to assess changes over time.
Results: In acute infarction, MR imaging estimates of function, perfusion, and viability showed no difference between the groups. In chronic infarction, however, VM202 increased maximum signal intensity and upslope at first-pass perfusion imaging and reduced infarct size at perfusion and delayed-enhancement imaging. These changes were associated with a decrease in end-diastolic (2.15 mL/kg ± 0.12 to 1.73 mL/kg ± 0.10, P < .01) and end-systolic (1.33 mL/kg ± 0.07 to 0.92 mL/kg ± 0.08, P < .001) volumes and an increase in ejection fraction (38.2% ± 1.3 to 47.0% ± 1.8, P < .001). In contrast, LV function deteriorated further in control animals. Compared with control animals, VM202-treated animals revealed peninsulas and/or islands of viable myocardium in infarcted and periinfarcted regions and greater number of capillaries (218 per square millimeter ± 19 vs 119 per square millimeter ± 17, P < .05) and arterioles (21 per square millimeter ± 4 vs 3 per square millimeter ± 1, P < .001).
Conclusion: Intramyocardial transfer of VM202 improved myocardial perfusion, viability, and LV function.
© RSNA, 2008
]]>Materials and Methods: The institutional review board approved this retrospective study, with waiver of informed consent. Included were 108 patients (71 men, 37 women; mean age, 61 years ± 8 [standard deviation]) with UIP (n = 79; 60 men, 19 women; mean age, 63 years ± 7.4) and fibrotic NSIP (n = 29; 11 men, 18 women; mean age, 57 years ± 12.9). Patients underwent pulmonary function tests (PFTs), bronchoalveolar lavage (BAL) fluid analysis, and thin-section CT. Two chest radiologists independently assigned scores for the extent of lung abnormalities detected at CT twice at 3-month intervals. The effect of histopathologic diagnoses and clinical and thin-section CT features on survival was evaluated by using Cox regression analyses.
Results: The 5-year survival rate (mean follow-up, 45 months) of patients with fibrotic NSIP was 76% in contrast to 46% for patients with UIP (P = .006). With multivariate analysis, a high fibrotic score (the extent of reticulation plus honeycombing) (hazard ratio = 1.200, P = .043) and an initial low diffusing capacity of lung for carbon monoxide (D
Conclusion: Patients with UIP or fibrotic NSIP who have a high fibrotic score determined at thin-section CT and a low D
© RSNA, 2008
]]>