© RSNA, 2009
]]>Materials and Methods: This study was approved by the institutional review board and was compliant with HIPAA; each patient gave informed consent. Thirty-five patients (mean age, 48 years ± 11 [standard deviation]; range, 29–75 years) with breast cancer were included. Treatment included doxorubicin and cyclophosphamide followed by a taxane-based regimen. Changes in tCho and tumor size in pCR versus non-pCR groups were compared by using the two-way Mann-Whitney nonparametric test. Receiver operating characteristic (ROC) analysis was performed to differentiate between them and the area under the ROC curve (AUC) was compared.
Results: In the pCR group, the tCho level change was greater compared with change in tumor size (P = .003 at first follow-up, P = .01 at second follow-up), but they were not significantly different in the non-pCR group. Changes in tumor size and tCho level at the first follow-up study were not significantly different between the pCR and non-pCR groups but reached significance at the second follow-up. In ROC analysis, the magnetic resonance (MR) imaging and MR spectroscopic parameters had AUCs of 0.65–0.68 at first follow-up; at second follow-up, AUC for change in tumor size was 0.9, AUC for change in tCho was 0.73.
Conclusion: Patients who show greater reduction in tCho compared with changes in tumor size are more likely to achieve pCR. The change in tumor size halfway through therapy was the most accurate predictor of pCR.
Supplemental material:
© RSNA, 2009
]]>Materials and Methods: The institutional review board exempted this HIPAA-compliant retrospective study from requiring informed consent. Structured reports from 48 744 consecutive pooled screening and diagnostic mammography examinations in 18 269 patients from April 5, 1999 to February 9, 2004 were collected. Mammographic findings were matched with a state cancer registry, which served as the reference standard. By using 10-fold cross validation, the Bayesian network was tested and trained to estimate breast cancer risk by using demographic risk factors (age, family and personal history of breast cancer, and use of hormone replacement therapy) and mammographic findings recorded in the Breast Imaging Reporting and Data System lexicon. The performance of radiologists compared with the Bayesian network was evaluated by using area under the receiver operating characteristic curve (AUC), sensitivity, and specificity.
Results: The Bayesian network significantly exceeded the performance of interpreting radiologists in terms of AUC (0.960 vs 0.939, P = .002), sensitivity (90.0% vs 85.3%, P < .001), and specificity (93.0% vs 88.1%, P < .001).
Conclusion: On the basis of prospectively collected variables, the evaluated Bayesian network can predict the probability of breast cancer and exceed interpreting radiologist performance. Bayesian networks may help radiologists improve mammographic interpretation.
Supplemental material:
© RSNA, 2009
]]>Materials and Methods: A breast tomosynthesis system was modeled by using a Monte Carlo program and a voxelized anthropomorphic breast phantom. The Monte Carlo program created simulated tomosynthesis projection images, which were reconstructed by using filtered back-projection software. Reconstructed images were analyzed for mass and microcalcification conspicuity, or the ratio of the lesion contrast to the anatomic and quantum noise surrounding the lesion. This analysis was performed at two compression levels (standard and 12.5% reduction) and for two breast compression thicknesses (4 and 6 cm). The change in conspicuity was analyzed for significance by using a bootstrap method and a paired Student t test.
Results: While keeping the glandular radiation dose constant with respective standard and reduced compression levels, the mean mass conspicuities were 1.39 ± 0.15 (standard error of the mean) and 1.46 ± 0.22 for a 4-cm breast compression phantom and 1.26 ± 0.15 and 1.22 ± 0.20 for a 6-cm breast phantom, and the mean microcalcification conspicuities were 16.2 ± 2.87 and 18.6 ± 2.63 for a 4-cm breast phantom and 11.4 ± 1.11 and 10.6 ± 1.18 for a 6-cm breast compression phantom.
Conclusion: For constant glandular dose, mass and microcalcification conspicuity remained approximately constant with decreased compression. Constant conspicuity implies that reduced compression would have a minimal effect on radiologists' performance, which suggests that there is justification for a measured reduction of breast compression for breast tomosynthesis, increasing the comfort of women undergoing the examination.
© RSNA, 2009
]]>Materials and Methods: This study was institutional review board approved, and all patients gave written informed consent. Fifty-five consecutive patients (36 men, 19 women; mean age, 61 years) who had recently experienced a stroke, had high-risk factors for thrombus formation, and underwent both two-phase 64-section cardiac CT angiography and TEE up to 5 days apart were examined. Agreement between CT and TEE for detection of thrombus was assessed with statistics. For quantitative analysis, the LAA–ascending aorta attenuation ratio (LAA/AA, in Hounsfield units) was measured on early- and late-phase CT images. The significance of differences in CT attenuation measurements were assessed by using the Student t test.
Results: A total of 14 thrombi were detected in the 55 patients at TEE. With TEE as the reference standard, the overall sensitivity, specificity, and positive and negative predictive values of cardiac CT angiography for the detection of thrombus in the LAA were 100% (14 of 14 patients), 98% (40 of 41 patients), 93% (14 of 15 patients), and 100% (40 of 40 patients), respectively. Concordance between cardiac CT angiography and TEE for the detection of thrombus in the LAA was high (overall = 0.953). Mean LAA/AA values were significantly different between thrombus (0.29 HU ± 0.12 [standard deviation]) and circulatory stasis (0.85 HU ± 0.12) on late-phase CT images (P < .001).
Conclusion: Two-phase 64-section cardiac CT angiography is a noninvasive sensitive modality for detecting LAA thrombi and differentiating thrombus from circulatory stasis in stroke patients.
© RSNA, 2009
]]>Materials and Methods: Institutional research ethics board approval was obtained. Following verbal consent, a telephone survey of radiologists working in Canada's 13 English-speaking and 13 U.S. university–affiliated radiology departments was performed. Participants were selected by using a multistage sampling scheme and simple random sampling within departments. Given a severe contrast material–induced allergic reaction case scenario, radiologists were first asked their initial medication of choice, then questioned specifically on the use of epinephrine. The Canadian and U.S. cohorts were compared by using the 2 and Fisher exact tests, as appropriate, and proportions and 95% confidence intervals (CIs) were computed.
Results: A total of 253 (81%) of 311 radiologists from a 30% target population were surveyed. Ninety-one percent (231 of 253; 95% CI: 88%, 94%) of radiologists chose epinephrine as the most important initial medication. No radiologist gave the ideal response, but 41% (94 of 231; 95% CI: 35%, 47%) provided an acceptable administration route, concentration, and dose; 17% (n = 39; 95% CI: 12%, 22%) of radiologists provided an overdose. Only 11% (27 of 253; 95% CI: 7%, 15%) of radiologists knew what concentration of epinephrine was available in their drug kit and/or crash cart and what equipment would be required to administer it to a patient.
Conclusion: Radiologists' knowledge of epinephrine for the management of severe contrast material–induced allergic reactions is deficient.
© RSNA, 2009
]]>Materials and Methods: Animal handling conformed to the Animal Care Committee Guidelines of all participating institutions. Single specimens of DVR were isolated from rats and perfused with a buffered solution containing iodixanol. A concentration of 23 mg of iodine per milliliter was chosen to mimic that expected to be used in usual examinations in humans. Luminal diameter was determined by using video microscopy, and NO was measured by using fluorescent techniques.
Results: Iodixanol led to 52% reduction of DVR luminal diameter, a narrowing that might interfere with passage of erythrocytes in vivo. Vasoconstriction induced by angiotensin II was enhanced by iodixanol. Moreover, iodixanol decreased NO bioavailability by more than 82%. Use of 4-hydroxy-2,2,6,6-tetramethylpiperidine 1-oxyl (a superoxide dismutase mimetic) prevented both vasoconstriction with iodixanol alone and increased constriction with angiotensin II caused by CM.
Conclusion: Iodixanol in doses typically used for coronary interventions constricts DVR, intensifies angiotensin II–induced constriction, and reduces bioavailability of NO. CM-induced nephropathy may be related to these events and scavenging of reactive oxygen species might exert a therapeutic benefit by preventing the adverse effects that a CM has on medullary perfusion.
Supplemental material:
© RSNA, 2009
]]>Materials and Methods: All experiments were approved by an institutional animal care and use committee. A total of 18 ablations were performed in vivo in normal porcine lungs. By using computed tomographic (CT) fluoroscopic guidance, a 17-gauge cooled triaxial microwave antenna (n = 9) and a 17-gauge cooled radiofrequency (RF) electrode (n = 9) were placed percutaneously. Ablations were performed for 10 minutes by using either 125 W of microwave power or 200 W of RF power delivered with an impedance-based pulsing algorithm. CT images were acquired every minute during ablation to monitor growth. Animals were sacrificed after the procedure. Ablation zones were then excised and sectioned transverse to the applicator in 5-mm increments. Minimum and maximum diameter, cross-sectional area, length, and circularity were measured from gross specimens and CT images. Comparisons of each measurement were performed by using a mixed-effects model; P < .05 was considered to indicate a significant difference.
Results: Mean diameter (3.32 cm ± 0.19 [standard deviation] vs 2.70 cm ± 0.23, P < .001) was 25% larger with microwave ablation and mean cross-sectional area (8.25 cm2 ± 0.92 vs 5.45 cm2 ± 1.14, P < .001) was 50% larger with microwave ablation, compared with RF ablation. With microwave ablation, the zones of ablation were also significantly more circular in cross section (mean circularity, 0.90 ± 0.06 vs 0.82 ± 0.09; P < .05). One small pneumothorax was noted during RF ablation but stabilized without intervention.
Conclusion: Microwave ablation with a 17-gauge high-power triaxial antenna creates larger and more circular zones of ablation than does a similarly sized RF applicator in a preclinical animal model. Microwave ablation may be a more effective treatment of lung tumors.
© RSNA, 2009
]]>Materials and Methods: All animal protocols and experiments were approved by the institutional animal care and use committee prior to study initiation. After injection of 0.2 or 0.4 mL of iodinated contrast material, six rats with mammary adenocarcinoma (three tumors each) were scanned in the axial mode for 5 minutes with 1-second ISD (reference scan), 2.5-mm section thickness, 2.5-mm interval, pitch of 1.3, 120 kV, 240 mA, and 0.5-second rotation time. A total of 126 dynamic data sets were created with commercial software by varying TST and ISD, including or excluding the IPVE, and using the aorta or inferior vena cava (IVC) as the input function. Comparative analyses were used to test for significant differences (t test, Wilcoxon signed rank test). Regression analysis was performed to assess the relationship between attenuation of the input function vessel and BF.
Results: No significant difference was observed (P > .05) when TST was as short as 30 seconds (range, 20–23 mL/100 g). In sequences performed with an ISD longer than 8 seconds, BF was significantly elevated (P < .01). Inclusion of the IPVE eliminated this difference (P > .10). Use of the IVC as the input function resulted in significantly higher BF (P < .02), with a correlation between peak attenuation and BF (R2 = 0.43).
Conclusion: To reduce radiation dose in tumor perfusion with CT, TST can be reduced without causing significant changes in BF calculation in an animal model. Scanning the aortic reference with peak contrast enhancement reduces variability sufficiently to allow for longer ISDs.
© RSNA, 2009
]]>Materials and Methods: The Animal Care and Use Committee approved this study. FN13762 murine breast cancer cells were implanted in 14 female Fischer 344 rats, and first-pass perfusion CT was performed. CT perfusion maps depicting blood flow, blood volume, mean transit time (MTT), and permeability–surface area (PSA) product were generated by using commercial perfusion software. The entire tumor area was divided into six separate regions on perfusion maps, and the regional perfusion indexes were quantified. Histologic vascular parameters, including microvessel density (MVD), luminal vessel number, luminal vessel area, and luminal vessel perimeter, were measured in the histologic region corresponding to the perfusion maps. Correlation analysis was performed between regional tumor perfusion indexes and histologic vascular parameters of the corresponding tumor region. Additionally, mean perfusion values of the entire tumor were correlated with histologic vascular parameters of the hot spot within the tumor. Among 14 rats, four were excluded from the analysis, and results were based on a final total of 10 rats.
Results: In tumors, blood flow, blood volume, and PSA product were significantly higher and MTT was significantly shorter (P < .05 for all) than these values in normal neck muscles. At region-by-region correlation, regional blood flow (r = 0.476), blood volume (r = 0.348), and MTT (r = –0.506) were significantly correlated with MVD in the corresponding tumor region (P < .01 for all). After adjustment for biologic variability between rats, regional blood flow (r = 0.614), blood volume (r = 0.515), MTT (r = –0.524), and PSA product (r = 0.228) remained significantly correlated with MVD in the corresponding tumor region. Correlation analysis between CT perfusion indexes of the entire tumor and histologic vascular parameters of the hot spot did not show significant correlations (P > .05).
Conclusion: Regional blood flow, blood volume, and MTT are significantly correlated with MVD in the corresponding tumor region.
© RSNA, 2009
]]>Materials and Methods: Protocol was approved by an institutional animal care and use committee. Caki-1, A498, and 786-0 human renal cell carcinoma (RCC) xenografts were implanted in 39 nude mice. Animals received 80 mg sorafenib per kilogram of body weight once daily once tumors measured 12 mm. ASL imaging was performed at baseline and day 14, with additional imaging performed for 786-0 and A498 (3 days to 12 weeks). Mean blood flow values and qualitative differences in spatial distribution of blood flow were analyzed and compared with histopathologic findings for viability and microvascular density. t Tests were used to compare differences in mean tumor blood flow. Bonferroni-adjusted P values less than .05 denoted significant differences.
Results: Baseline blood flow was 80.1 mL/100 g/min ± 23.3 (standard deviation) for A498, 75.1 mL/100 g/min ± 28.6 for 786-0, and 10.2 mL/100 g/min ± 9.0 for Caki-1. Treated Caki-1 showed no significant change (14.9 mL/100 g/min ± 7.6) in flow, whereas flow decreased in all treated A498 on day 14 (47.9 mL/100 g/min ± 21.1) and in 786-0 on day 3 (20.3 mL/100 g/min ± 8.7) (P = .003 and .03, respectively). For A498, lowest values were measured at 28–42 days of receiving sorafenib. Regions of increased flow occurred on days 35–49, 17–32 days before documented tumor growth and before significant increases in mean flow (day 77). Although 786-0 showed new, progressive regions with signal intensity detected as early as day 5 that correlated to viable tumor at histopathologic examination, no significant changes in mean flow were noted when day 3 was compared with all subsequent days (P > .99).
Conclusion: ASL imaging provides clinically relevant information regarding tumor viability in RCC lines that respond to sorafenib.
© RSNA, 2009
]]>Materials and Methods: The study protocol was approved by the human research committee, and all patients gave written informed consent. One hundred fifty patients (83 male, 67 female; mean age, 42.6 years; age range, 17–84 years) who were clinically suspected of having small-bowel neoplasm and whose previous upper and lower gastrointestinal endoscopy findings were normal underwent MR enteroclysis. The MR enteroclysis findings were prospectively evaluated for the presence of focal bowel wall thickening, small-bowel masses, and small-bowel stenosis. Positive MR enteroclysis findings were compared with histopathologic examination results obtained after surgical (n = 19) or endoscopic (n = 2) procedures. Negative MR enteroclysis results were compared with the results of enteroscopy (n = 5), capsule endoscopy (n = 53), or conventional enteroclysis with subsequent clinical follow-up (n = 71). The diagnostic performance of MR enteroclysis was analyzed on a per-patient basis.
Results: MR enteroclysis was successfully completed in all 150 patients and enabled correct detection of 19 small-bowel neoplasms, which were confirmed at histopathologic examination: three carcinoid neoplasms, two adenocarcinomas, two stromal tumors, five lymphomas, one angiomatous mass, three small-bowel metastases, one leiomyoma, one adenoma, and one lipoma. Overall sensitivity, specificity, and accuracy in identifying patients with small-bowel lesions were 86% (19 of 22), 98% (126 of 128), and 97% (145 of 150), respectively. Two MR enteroclysis examinations yielded false-positive findings, and three yielded false-negative findings.
Conclusion: MR enteroclysis is an accurate modality for detecting small-bowel neoplasms in symptomatic patients.
Supplemental material:
© RSNA, 2009
]]>Materials and Methods: The institutional review board approved the study protocol; informed consent was obtained from all participants. Thirty-one consecutive patients who had CD or who were suspected of having CD underwent CT and MR enterography, SBFT, and ileocolonoscopy. Two independent readers reviewed CT and MR enterographic and SBFT images for presence of active terminal ileitis and extraenteric complications. Accuracy values of CT and MR enterography and SBFT for identification of active terminal ileitis were evaluated with the receiver operating characteristic method, with ileocolonoscopic findings as the reference standard. Sensitivity values of CT and MR enterography and SBFT for detection of extraenteric complications were compared by using the McNemar test, with results of imaging studies, surgery, and physical examination as reference standards.
Results: The study population included 30 patients (17 men, 13 women; mean age, 29.0 years) with CD. Differences in areas under the receiver operating characteristic curves for CT enterography (0.900 and 0.894), MR enterography (0.933 and 0.950), and SBFT (0.883 and 0.928) for readers 1 and 2, respectively, in the detection of active terminal ileitis were not significant (P > .017). Sensitivity values for detection of extraenteric complications were significantly higher for CT and MR enterography (100% for both) than they were for SBFT (32% for reader 1 and 37% for reader 2) (P < .001).
Conclusion: Because MR enterography has a diagnostic effectiveness comparable to that of CT enterography, this technique has potential to be used as a radiation-free alternative for evaluation of patients with CD.
Supplemental material:
© RSNA, 2009
]]>Materials and Methods: The institutional review board approved this HIPAA-compliant retrospective study; informed consent was waived. Radiologic database review revealed 1655 SBFT studies over 6 years in 1282 patients after RYGBP. IH was suggested on 24 studies in 23 patients. Studies were analyzed for atypical bowel configuration, change in bowel or suture position, and obstruction. Chart review was performed to determine clinical course, treatment, and outcome. Studies from a control group of 21 RYGBP patients were similarly analyzed. Statistical comparison was performed with the Fisher exact test.
Results: Clinical and/or surgical evidence of IH was found following 21 SBFT studies in 20 of 1282 patients (1.6%). Atypical bowel configuration with clustered small bowel was identified on all studies. Cluster location was lateral to descending colon (n = 10), left upper quadrant (n = 6), left upper and mid abdomen (n = 3), right midabdomen (n = 2), under the gastric pouch (n = 1), and right lower quadrant (n = 1). For two studies, two locations of clustered bowel were identified. Change in jejunojejunal suture position occurred in all cases with radiopaque suture (n = 15). Other signs of IH included displaced colon (n = 19), visible entrance and exit limbs into the hernia (n = 17), stasis in clustered bowel (n = 16), densely matted bowel (n = 12), and a straight left lateral border of clustered bowel (n = 10). Partial obstruction occurred in 16 patients. Findings of atypical bowel configuration, clustered bowel, and staple line change were significant when compared with the control.
Conclusion: IH following RYGBP is a rare but potentially fatal complication. Radiologists must be aware of this complication and its diagnostic features at SBFT.
Supplemental material:
© RSNA, 2009
]]>Materials and Methods: This prospective single-center HIPAA-compliant study had institutional review board approval, and written informed consent was obtained. Forty-eight patients (31 men, 17 women; age range, 35–77 years) with 60 malignant hypervascular liver tumors (mean diameter, 20.1 mm ± 16.4 [standard deviation]) were enrolled. Pathologic proof of focal lesions was obtained with histopathologic analysis for 33 nodules and imaging follow-up after a minimum of 12 months for 27 nodules. Patients underwent dual-energy 64-section multi–detector row CT. By using vendor-specific software, two imaging protocols—140 kVp and 385 mA (protocol A) and 80 kVp and 675 mA (protocol B)—were compared during the late hepatic arterial phase of contrast enhancement. Paired t tests were used to compare tumor-to-liver contrast-to-noise ratio (CNR) for each lesion, mean image noise, and effective dose between the two data sets. Three readers qualitatively assessed the two data sets in a blinded and independent fashion. Lesion detection and characterization and reader confidence were recorded, as well as readers' subjective evaluations of image quality. Wilcoxon-Mann-Whitney statistical analysis was performed on this assessment.
Results: Image noise increased from 5.7 to 11.4 HU as the tube voltage decreased from 140 to 80 kVp (P < .0001), resulting in a significantly lower image quality score (4.0 vs 3.0, respectively) with protocol B according to all readers (P < .001). At the same time, protocol B yielded significantly higher CNR (8.2 vs 6.4) and lesion conspicuity scores (4.6 vs 4.1) than protocol A, along with a lower effective dose (5.1 vs 17.5 mSv) (P < .001 for all).
Conclusion: By substantially increasing the tumor-to-liver CNR, a low tube voltage, high tube current CT technique improves the conspicuity of malignant hypervascular liver tumors during the late hepatic arterial phase while significantly reducing patient radiation dose.
Supplemental material:
© RSNA, 2009
]]>Materials and Methods: The institutional review board approved the study and waived the informed consent requirement. A total of 259 patients, 37 (14.3%) of whom had previously undergone appendectomy, underwent LD unenhanced CT (mean effective dose, 1.7 mSv) performed with a 16– or 64–detector row scanner to assess urinary colic. Three readers used the sliding slab averaging technique to retrospectively review the thin-section (0.67- or 2.00-mm section thickness) images and grade the appendix as absent, unsurely or partly visualized, or clearly and entirely visualized. Interobserver agreement was measured with weighted statistics. McNemar tests were used to compare sensitivity between the readers. Logistic regression analysis was performed to assess the effects of body mass index, patient sex, and type of CT scanner on appendiceal visualization.
Results: The statistics for each reader pair were as follows: 0.97 for agreement between readers 1 and 2, 0.93 for agreement between readers 2 and 3, and 0.92 for agreement between readers 1 and 3. Each reader clearly identified the entire appendix in 213 (96.0%), 209 (94.1%), and 205 (92.3%) of the 222 patients without a history of appendectomy. When unsurely or partly visualized appendices were included, the frequencies increased to 99.1% (n = 220), 98.7% (n = 219), and 97.3% (n = 216), respectively, for readers 1, 2, and 3. These frequencies rarely differed between the readers. (P values ranged from .021 to greater than .99.) The three readers consistently reported that the appendix was not visualized in the 37 patients who had undergone appendectomy. None of the tested variables significantly affected appendix visualization.
Conclusion: Most normal appendices are visualized on thin-section LD unenhanced CT images reviewed with the sliding slab averaging technique.
Supplemental material:
© RSNA, 2009
]]>Materials and Methods: The study was approved by the institutional review board. Informed consent was obtained. The prospective study included 20 consecutive prospective patients (age range, 37–56 years) for whom preprocedural prediction of the best tube angle obliquity was determined by using 3D-reconstructed contrast-enhanced MR angiography; the best tube angle obliquity was provided to the interventionist. Three-dimensional reconstruction was performed by using an application of the angiographic unit. The radiation dose, fluoroscopy time, and contrast medium volume for those patients were compared with those data in 20 retrospectively assessed control patients (age range, 39–56 years) from the prior 20 procedures performed by the same interventionist.
Results: Tube angle prediction resulted in a significant reduction in the radiation dose utilized (P < .001), fluoroscopy time (P = .002), and contrast medium volume (P < .001) for the sample patients compared with those for the control patients. Overall radiation dose was reduced from a mean of 11 044 µGy per square meter to a mean of 4172.5 µGy per square meter. Fluoroscopy time was reduced from a mean of 15 minutes 30 seconds to 8 minutes 49 seconds. Contrast medium volume was reduced from a mean of 135 mL to 75 mL.
Conclusion: Preprocedural prediction of the best tube angle obliquity for visualization of the origin of the uterine artery by using 3D-reconstructed contrast-enhanced MR angiography results in significant reductions in radiation dose, fluoroscopy time, and contrast medium volume during UAE.
© RSNA, 2009
]]>Materials and Methods: Institutional review board approval was obtained for this HIPAA-compliant retrospective study, and the informed consent requirement was waived. Two independent observers retrospectively assessed MR images obtained in 41 patients with non–fat-containing T1 hyperintense renal lesions. The MR examination included acquisition of DW and contrast-enhanced T1-weighted images. For each index lesion, the observers assessed the (a) mean (± standard deviation) of ADC, (b) enhancement ratio, and (c) subtracted images for the presence of enhancement (confidence score, 1–5). Histopathologic analysis of renal cell carcinomas (RCCs) and follow-up imaging for benign lesions were the reference standards. ADCs of benign lesions and RCCs were compared. Receiver operating characteristic (ROC) curve analysis was performed to assess the accuracy of DW imaging, enhancement ratio, and subtraction for the diagnosis of RCC.
Results: A total of 64 lesions (mean diameter, 3.9 cm), including 38 benign T1 hyperintense cysts and 26 RCCs, were assessed. Mean ADCs of RCCs were significantly lower than those of benign cysts ([1.75 ± 0.57] x 10–3 mm2/sec vs [2.50 ± 0.53] x 10–3 mm2/sec, P < .0001). ADCs of solid and cystic portions of complex cystic RCCs were significantly different ([1.37 ± 0.55] x 10–3 mm2/sec vs [2.45 ± 0.63] x 10–3 mm2/sec, P < .0001). When data from both observers were pooled, area under the ROC curve, sensitivity, and specificity were 0.846, 71%, and 91%, respectively, for DW imaging; 0.865, 65%, and 96%, respectively, for enhancement ratio (at the excretory phase); and 0.861, 83%, and 89%, respectively, for subtraction (P = .48 and P = .85, respectively). The combination of DW imaging and subtraction resulted in area under the ROC curve, sensitivity, and specificity of 0.893, 87%, and 92%, respectively, with significantly improved reader confidence compared with subtraction alone (P = .041).
Conclusion: The performance of DW imaging was equivalent to that of enhancement ratio in the characterization of T1 hyperintense renal lesions, with both methods having lower sensitivity than image subtraction without reaching significance.
Supplemental material:
© RSNA, 2009
]]>© RSNA, 2009
]]>Materials and Methods: This study was performed according to a protocol approved by the institutional animal care committee. A Monte Carlo simulation was used to determine the pair of near-infrared (NIR) dyes that was best suited for the normalization algorithm. The authors tested the correction method in vitro and in vivo by injecting free dye mixtures intramurally in the aortas of four rats. The potential clinical utility was then evaluated by applying the method to the endovascular measurement of protease activity in a rat model of focal aortic injury.
Results: When the Monte Carlo simulation was used in the normalization algorithm, it was predicted that the intensities of signals from two NIR dyes would vary ±3% across 1 mm of blood compared with the intensity of the raw fluorochrome signal, which would vary ±60%. This result was validated in vitro. Endovascular imaging of free dye collections revealed that clinically relevant, uncontrollable differences in the amount of blood intervening between the imaging catheter and the dye collection precipitated dramatic variations in raw NIR fluorescence. However, use of the correction method resolved these variations such that the measured signal intensity correlated well with the different dye concentrations in the different animals. Moreover, endovascular imaging of the focal aortic injury model enabled successful measurement of enzyme activity in the walls of the rat aortas.
Conclusion: The authors implemented a correction method for quantitative real-time endovascular imaging of fluorescence that enables one to resolve the attenuating effects of blood on NIR signal.
© RSNA, 2009
]]>Materials and Methods: This retrospective study was approved by the institutional ethics review board; informed consent was obtained from the patients. A total of 250 consecutive patients (mean age, 62.0 years; range, 25–89 years) consisting of 190 women (mean age, 64 years; range, 25–89 years) and 60 men (mean age, 57.0 years; range, 27–83 years) who were suspected of having osteoporosis and who underwent VFA in the supine position and radiography of the spine were evaluated. VFA and semiquantitative radiography were analyzed by using a six-marker point method to describe the shape and deformity of each vertebra. Visual radiography of the lateral spine was performed by an experienced radiologist. The agreement between VFA, visual radiography, and semiquantitative radiography of semiquantitative graded fractures was assessed by using weighted statistics.
Results: Visual radiography helped identify 92 (36.8%) patients with at least one vertebral fracture (mean, 1.8 per patient). Most fractures were present in T7, T12, and L1. Excellent agreement was found between VFA and visual radiography, with 97.5% concordance and = 0.82; VFA and semiquantitative radiography were in agreement in 97.4% of patients, with = 0.83; and visual radiography and semiquantitative radiography were in agreement in 98.1%, with = 0.87. Sensitivity, specificity, and positive and negative predictive values calculated by lesion level for VFA compared with visual assessment were 83.6%, 99.1%, 84.1%, and 99.1%, respectively.
Conclusion: VFA performed with patients in the supine position is an accurate method to help detect vertebral fractures when compared with conventional spine radiography. VFA permits combination of fracture assessment with bone mineral density measurement in a single session.
© RSNA, 2009
]]>Materials and Methods: Institutional ethics approval and written informed consent were obtained. Thirty-two patients (30 women, two men; mean age, 52 years) with RA twice underwent MR of either the most symptomatic hand (n = 21) or wrist (n = 11), once performed with a 1.0-T extremity MR system and once with a 1.5-T conventional MR system. The MR examinations were independently assessed by two radiologists blinded to imaging platform and patient clinical information for erosions, synovitis, and bone marrow edema (BME), according to the Rheumatoid Arthritis MR Imaging Score (RAMRIS). One radiologist reevaluated all cases a second time to determine the intraobserver variability for each system. Patient comfort was assessed with a questionnaire. Intraclass correlation coefficients (ICCs) and smallest detectable differences (SDDs) were measured.
Results: ICCs for intermachine agreement were 0.97–0.99 for erosions, 0.88–0.97 for synovitis, and 0.98–0.99 for BME for both readers. The SDDs between the two systems, expressed as a percentage of the maximum score, ranged from 3.3% to 12.2% for erosions, from 7.4% to 14.8% for synovitis, and from 5% to 9.9% for BME for both readers. ICCs for interreader agreement ranged from 0.69 to 0.99 and for intrareader agreement, from 0.88 to 0.99. There were substantial differences in the subjective patient assessment of confinement, system noise, and difficulty with immobilization, and 95.8% of patients preferred examinations performed with extremity MR.
Conclusion: The 1.0-T extremity MR system demonstrates synovial and osseous changes in RA equally as well as a 1.5-T conventional MR system and is preferred by patients.
© RSNA, 2009
]]>Materials and Methods: The hospital's institutional review board approved this retrospective study and waived informed consent. Fifty patients (23 male patients, 27 female patients; median age, 53 years) with newly diagnosed astrocytic tumors were evaluated. Two observers blinded to clinical information independently measured the ADCs by manually placing three to five regions of interest (40–60 mm2) within the solid tumor either with or without contrast material–enhanced components and calculated the average ADC. Minimum and maximum ADCs were selected, and the difference between them was recorded as the ADC difference value. These ADC values were used as the parameters for tumor grading and were compared by using the Kruskal-Wallis test and receiver operating characteristic (ROC) curve analysis.
Results: According to ROC analyses for distinguishing tumor grade, minimum ADCs showed the largest areas under the ROC curve. Minimum ADCs optimally helped distinguish grade 1 from higher-grade tumors at a cutoff value of 1.47 x 10–3 mm2/sec and grade 4 from lower-grade tumors at a cutoff value of 1.01 x 10–3 mm2/sec (P < .001 for both). ADC difference values helped distinguish grade 2 from grade 3 tumors at a cutoff value of 0.31 x 10–3 mm2/sec (P < .001). When tumors were graded by using the combined minimum ADC and ADC difference cutoff values mentioned above (the two-parameter method), the following positive predictive values were obtained: grade 1 tumors, 73% (eight of 11); grade 2 tumors, 100% (five of five); grade 3 tumors, 67% (eight of 12); and grade 4 tumors, 91% (20 of 22).
Conclusion: Using a combination of minimum ADCs and ADC difference values (the two-parameter method) facilitates the accurate grading of astrocytic tumors.
© RSNA, 2009
]]>Materials and Methods: The medical ethics committee approved the ATENA study, and all patients gave informed consent for participation in the study. The ATENA study was performed in 27 institutions. For each patient group, we recorded aneurysm characteristics, rate of adverse events related to the treatment, and patient outcome.
Results: In this study, 547 patients (383 women, 164 men; mean age, 51.0 years ± 11.1 [standard deviation]; range, 22–83 years) with 572 aneurysms were included; 325 patients were treated with coils alone and 222 patients were treated with the remodeling technique. The overall rate of adverse events related to the treatment—regardless of whether the adverse events led to clinical consequences—was 10.8% (35 of 325) for treatment with coils alone and 11.7% (26 of 222) for the remodeling technique. Thromboembolic events, intraoperative rupture, and device-related problems were encountered in 20 (6.2%), seven (2.2%), and eight (2.5%) of 325 patients in the standard treatment group and in 12 (5.4%), seven (3.2%), and seven (3.2%) of 222 patients in the remodeling technique group, respectively. The morbidity and mortality rates did not differ significantly between groups: 2.2% (seven of 325) and 0.9% (three of 325) in the standard treatment group and 2.3% (five of 222) and 1.4% (three of 222) in the remodeling technique group, respectively.
Conclusion: The remodeling technique was associated with a similar rate of adverse events and morbidity and mortality combined compared with the standard treatment with coils, and, thus, the remodeling technique is as safe as the standard treatment with coils.
© RSNA, 2009
]]>Materials and Methods: MR images in 56 patients (30 male, 26 female; mean age ± standard deviation, 11.9 years ± 8.5) with genetically confirmed L2HGA were retrospectively reviewed, with institutional review board approval and waiver of informed consent. At least one complete series of transverse T2-weighted images was available for all patients. The images were evaluated by using a previously established scoring list. The correlation between MR imaging abnormalities and disease duration was assessed (Mann-Whitney or Kruskal-Wallis test).
Results: The cerebral white matter (WM) abnormalities preferentially affected the frontal and subcortical regions. The abnormal subcortical WM often had a mildly swollen appearance (37 patients). Initially, the WM abnormalities were at least partially multifocal (32 patients). In patients with longer disease duration, the WM abnormalities became more confluent and spread centripetally, but the periventricular rim remained relatively spared (41 patients). The mean disease duration in patients with WM atrophy (14.8 years) was significantly longer (P = .001) than that in patients without atrophy (6.7 years). Bilateral involvement of the globus pallidus (55 patients), caudate nucleus (56 patients), and putamen (56 patients) was seen at all stages. The cerebellar WM was never affected. The dentate nucleus was involved bilaterally in 55 of 56 patients.
Conclusion: L2HGA has a distinct highly characteristic pattern of MR imaging abnormalities: a combination of predominantly subcortical cerebral WM abnormalities and abnormalities of the dentate nucleus, globus pallidus, putamen, and caudate nucleus. With increasing disease duration, WM abnormalities and basal ganglia signal intensity abnormalities become more diffuse and cerebral WM atrophy ensues.
© RSNA, 2009
]]>Materials and Methods: Institutional review board approval and informed consent were obtained for this prospective study. Fifty-two patients (41 women, 11 men; mean age, 75.9 years; range, 56–95 years) were enrolled from August 2006 to August 2007. All patients underwent preoperative magnetic resonance (MR) imaging and assessment of mobility and pain scores. PV was performed and patients underwent 1-, 3-, and 6-month follow-up examinations. MR findings of fluid sign (a focal, linear, or triangular area of strong hyperintensity, which is isointense relative to cerebrospinal fluid on T2-weighted sagittal images) and vacuum cleft were analyzed with respect to clinical outcome. Data were analyzed by using a combination of independent Student t test, 2 test, analysis of variance, and Fisher exact test.
Results: Thirty-four (65%) patients showed vacuum cleft; 14 (27%) showed fluid sign at the bone-cement interface. Patients without fluid sign in the treated vertebral bodies had better mobility and pain improvement compared with patients with fluid sign at 1- and 3-month follow-up (P < .05). The adjacent fracture percentage (seven of 14, 50%) was higher in patients with fluid sign in the treated vertebral bodies than in those without (three of 38, 8%) (P = .002). Pain was similar in groups with and without fluid sign at 6-month follow-up.
Conclusion: The fluid sign in the treated vertebral bodies had a high negative predictive value of 92% and a positive predictive value of 50% to develop a new adjacent compression fracture.
© RSNA, 2009
]]>Materials and Methods: This retrospective study had institutional review board approval, and informed consent was waived. Fluid-attenuated inversion-recovery (FLAIR) MR images and T2-weighted MR images in the oblique coronal plane in 22 consecutive patients (10 men, 12 female patients; mean age, 41.0 years; range, 14–76 years) (25 hemispheres) with a clinical diagnosis of MTLE were retrospectively evaluated. Twenty-five hippocampi in 15 subjects without epilepsy were evaluated as age-matched controls. The volumes and thicknesses of the four anatomic sections of the hippocampi were quantitatively measured on the T2-weighted images. Two radiologists independently reviewed the MR imaging findings of the hippocampus regarding atrophy, abnormal signal intensity, and PLHS on each side separately, without comparing both sides. Sensitivity and specificity were calculated among the MR imaging findings.
Results: Signal intensity abnormality on FLAIR images had a sensitivity of 36%, a specificity of 96%, and an accuracy of 66% for the diagnosis of hippocampal sclerosis. PLHS on T2-weighted MR images had a sensitivity of 76% and a specificity of 80% for the diagnosis of hippocampal sclerosis. The sensitivity for PLHS was higher than that for atrophy (44%) and abnormal signal intensity (48%) of the hippocampus on T2-weighted MR images. Although the mean volume of the hippocampus and the thickness of the hippocampal body were significantly smaller for patients with MTLE than for control subjects (P < .001 for both), there was no clear distinguishing threshold value between abnormal and normal hippocampi.
Conclusion: PLHS showed the highest sensitivity for MTLE. This MR imaging feature might improve the accuracy of the diagnosis of bilateral hippocampal sclerosis, although further research is required.
© RSNA, 2009
]]>Materials and Methods: This was a prospective, HIPAA-compliant institutional review board–approved investigation with written informed parental consent. Nineteen English-speaking, right-handed children with a normal IQ and developmental dyslexia (16 male, three female; age range, 6–16 years; mean age, 9.9 years) and 18 normal-reading, age-matched pediatric control subjects (13 male, five female; age range, 6–15 years; mean age, 10.0 years) underwent DT imaging (30 directions, three signals acquired, voxel size of 2 mm). Regions of interest were placed on the SLF, IFO-ILF, and PLIC, and tensor metrics were calculated. Statistical analyses of differences in cognitive function between the dyslexic and control groups were performed by using the two-sample t test. Differences in tensor metrics were examined by using analysis of covariance models.
Results: In the control subjects, the fractional anisotropy (FA) of all tracts studied increased with age. In the dyslexic subjects, the age-related increases in FA in the SLF were most similar to those in the control subjects (P = .504), while mean FA values for the IFO-ILF (P = .009) and PLIC (P < .0001) were higher than those in the control subjects up to around 11 years of age, after which they were lower. Apparent diffusion coefficients consistently decreased in both groups. There was a nonsignificant increase in mean axial diffusivity in the IFO-ILF in the control group but not in the dyslexia group. Increases in axial diffusivity seen in the PLIC in the control group were not seen in the dyslexia group. There were no marked differences in tensor metrics between the left and right hemispheres within or between the two groups.
Conclusion: Findings at 3.0-T DT imaging suggest that white matter differences in dyslexic children are not limited to the portion of the brain traditionally considered to be integral to word recognition and processing.
© RSNA, 2009
]]>Materials and Methods: This HIPAA-compliant retrospective study was approved by the institutional review board, and informed consent was waived. The records of twelve patients with large intracranial tumors who underwent embolization were analyzed for imaging characteristics that would portend acute neurologic deterioration following embolization. The degree of midbrain compression was calculated by using the cisternal angle (the angle formed at the intersection of a line drawn along the midsagittal plane and a line drawn along the anterior aspect of the cerebral peduncle). Angiograms were evaluated for the degree of pre- and postembolization tumor blush. Neurologic status before and after embolization was evaluated. The Wilcoxon signed rank test was used to compare the cisternal angles ipsilateral and contralateral to the tumor. The cisternal angle was measured in 100 control subjects with no mass lesions to evaluate its normal distribution.
Results: Of the 12 patients, three experienced acute clinical deterioration after embolization. A feature common to these patients was substantial preprocedure midbrain compression, as indicated by a cisternal angle of less than 25°, which was significantly less than the mean angle in the control group. Another consistent risk factor was a strong initial tumor blush pattern and a major blush reduction following embolization.
Conclusion: Cisternal angle is an objective measure of midbrain compression. The presence of a cisternal angle less than 25° (indicating severe midbrain compression), strong tumor blush, and major postprocedure blush reduction are predictors of clinical deterioration after embolization.
© RSNA, 2009
]]>Supplemental material:
© RSNA, 2009
]]>Materials and Methods: Protocols approved by the National Institutes of Health Animal Care and Use Committee were used to examine ex vivo the hind limb of a rat and the toe of a pig. The Fourier x-ray scattering imaging technique involves the use of a grid mask to modulate the cone beam and Fourier spectral filters to isolate the harmonic images. The technique yields attenuation, scattering, and phase-contrast (PC) images from a single exposure. In the rat tibia cortical bone, the scattering signals from two orthogonal grid orientations were compared by using Wilcoxon signed rank tests. In the pig toe, the heterogeneity of scattering and PC signals was compared between trabecular and compact bone regions of uniform attenuation by using F tests.
Results: In cortical bone, the scattering signal was significantly higher (P < 10–15) when the grid was parallel to the periosteal surface. Trabecular bone, as compared with cortical bone, appeared highly heterogeneous on the scattering (P < 10–34) and PC (P < 10–27) images.
Conclusion: The ordered alignment of the mineralized collagen fibrils in compact bone was reflected in the anisotropic scattering signal in this bone. In trabecular bone, the porosity of the mineralized matrix accounted for the granular pattern seen on the scattering and PC images.
Supplemental material:
© RSNA, 2009
]]>Materials and Methods: This study was approved by an institutional review board. Written informed consent was obtained from all participants. Data were collected from the Multicentric Italian Lung Detection project, a randomized controlled lung cancer screening trial. The first 1236 consecutive baseline computed tomographic (CT) studies performed at the Istituto Nazionale Tumori of Milan were evaluated. Among the enrolled participants, those who underwent repeat low-dose CT after 3 months and had at least one indeterminate nodule with a volume of more than 60 mm3 (diameter of 4.8 mm or greater) were considered. Nonsolid, part-solid, and pleural-based nodules were excluded from this study. A descriptive analysis was performed by calculating means and standard deviations of nodule volumes at three assessment times (at baseline and 3 and 12 months later). The volume measurement repeatability was determined by using the approach described by Bland and Altman.
Results: One hundred one subjects (70 men, 31 women; mean age, 58 years) with 233 eligible nodules (mean volume, 98.3 mm3; range, 5–869 mm3) were identified. The 95% confidence interval for difference in measured volumes was in the range of ±27%. About 70% of measurements had a relative difference in nodule volume of less than 10%. No malignant lesions were registered during the follow-up of these subjects.
Conclusion: Semiautomatic volumetry is sufficiently accurate and repeatable and may be useful in assisting with lung nodule management in a lung cancer screening program.
© RSNA, 2009
]]>Materials and Methods: This retrospective study was institutional review board approved; 107 patients treated for symptomatic VM were evaluated. After informed consent was obtained, 66 patients were sent a questionnaire regarding treatment effectiveness and patient satisfaction. Patient files and imaging data were retrieved to obtain information regarding the VMs and VM treatment. Kaplan-Meier survival curves were constructed to analyze clinical success rates over time.
Results: The most frequent reasons for patients to seek treatment were pain (89%, n = 59) and swelling (91%, n = 60). The majority of VMs were the low-flow venous type (83%, n = 55). Three months after treatment, clinical success was reported for 58% (n = 38) of patients and clinical failure was reported for 42% (n = 28). At 1-, 2-, 3-, 4-, and 5-year follow-up, clinical success was 49%, 49%, 42%, 42%, and 42%, respectively. Twenty-seven (40%) patients experienced complications, 12 of which required additional treatment. In all, 35 (53%) patients reported being satisfied with their treatment. Patient satisfaction was closely correlated with clinically successful long-term outcome of treatment.
Conclusion: Initial partial or complete relief of VM complaints after percutaneous treatment is expected in 58% of patients, irrespective of VM size or classification. These results were durable over a 5-year follow-up period.
Supplemental material:
© RSNA, 2009
]]>Materials and Methods: The study was approved by the institutional review board; informed consent was waived because of the retrospective design. Over a 13-year period, 1136 patients with 1928 malignant liver tumors underwent ultrasonographically guided percutaneous MW ablation (583 with a noncooled-shaft antenna and 553 with a cooled-shaft antenna). A total of 3697 MW ablation sessions (average, 1.8 sessions per patient) were performed. Mortality and treatment-related major and minor complications were documented. Data were subsequently analyzed to determine whether the major complication rate was related to antenna type, tumor size, tumor location, or number of MW sessions.
Results: Two deaths not directly attributable to MW ablation were encountered. Major complications occurred in 30 (2.6%) patients and included liver abscess and empyema (n = 5), bile duct injury (n = 2), perforation of the colon (n = 2), tumor seeding (n = 5), pleural effusion requiring thoracentesis (n = 12), hemorrhage requiring arterial embolization (n = 1), and skin burn requiring resection (n = 3). Minor complications included fever; pain; asymptomatic pleural effusion, gallbladder wall thickening, and arterioportal shunt; small stricture of the bile duct; and skin burn requiring no treatment. Use of noncooled-shaft antenna and an increased number of MW ablation sessions were associated with a higher rate of major complications (P < .05).
Conclusion: MW ablation is a well-tolerated technique with an acceptably low rate of major complications for treatment of malignant liver tumors. Use of a cooled-shaft antenna, as well as fewer MW sessions, may help minimize major complications.
© RSNA, 2009
]]>