Organ-specific response to nivolumab in patients with non-small cell lung cancer (NSCLC).

BACKGROUND: Response to immune checkpoint inhibitors depends on tumor intrinsic properties and also on host factors in the tumour microenvironment including the presence of immune cells (IC). We hypothesized that nivolumab efficacy varies across different metastatic sites. METHODS: We retrospectively analyzed computed tomography scans of patients with metastatic non-small cell lung carcinoma (NSCLC) receiving nivolumab. RECIST 1.1 criteria were applied to assess the overall response rate (ORR) and organ-specific response rate (OSRR). RESULTS: We analyzed 52 patients including 44% females, 58% adenocarcinoma and 8% never smokers. Involved organs had target-lesions in the lung (42%), liver (25%), lymph nodes (56%) and soft tissue (13%) and non-target lesions in the bones (23%). ORR and disease control rate (DCR) were 20% and 45%, respectively. Median overall survival, progression-free survival and duration of response were 11.9, 2.3 and 10.3 months. OSRR and organ-specific DCR (OSDCR) were 28% and 90% in lymph nodes, 8% and 54 in the liver, and 9% and 55% in lung metastases. Nine out of 12 patients with bone metastases had progressive lesions. The cumulative incidence probability of organ-specific progression at 6 months was 14% in lymph nodes, 42% in the liver, 36% in lung metastases and 26% in the primary tumor, 29% in soft tissue and 33% in adrenal metastases. CONCLUSION: In conclusion, the efficacy of immunotherapy is dependent on the metastatic location. Treatment appears more active in lymph nodes compared to other organ sites such as liver, adrenals and bone. Future strategies may include additional local treatment in case of oligoprogression in these organs in patients with otherwise sustained treatment benefit.

Ultralow dose CT for pulmonary nodule detection with chest x-ray equivalent dose - a prospective intra-individual comparative study.

PURPOSE: To prospectively evaluate the accuracy of ultralow radiation dose CT of the chest with tin filtration at 100 kV for pulmonary nodule detection. MATERIALS AND METHODS: 202 consecutive patients undergoing clinically indicated chest CT (standard dose, 1.8 ± 0.7 mSv) were prospectively included and additionally scanned with an ultralow dose protocol (0.13 ± 0.01 mSv). Standard dose CT was read in consensus by two board-certified radiologists to determine the presence of lung nodules and served as standard of reference (SOR). Two radiologists assessed the presence of lung nodules and their locations on ultralow dose CT. Sensitivity and specificity of the ultralow dose protocol was compared against the SOR, including subgroup analyses of different nodule sizes and types. A mixed effects logistic regression was used to test for independent predictors for sensitivity of pulmonary nodule detection. RESULTS: 425 nodules (mean diameter 3.7 ± 2.9 mm) were found on SOR. Overall sensitivity for nodule detection by ultralow dose CT was 91%. In multivariate analysis, nodule type, size and patients BMI were independent predictors for sensitivity (p < 0.001). CONCLUSIONS: Ultralow dose chest CT at 100 kV with spectral shaping enables a high sensitivity for the detection of pulmonary nodules at exposure levels comparable to plain film chest X-ray. KEYPOINTS: • 91% of all lung nodules were detected with ultralow dose CT • Sensitivity for subsolid nodule detection is lower in ultralow dose CT (77.5%) • The mean effective radiation dose in 202 patients was 0.13 mSv • Ultralow dose CT seems to be feasible for lung cancer screening.

Noise Texture Deviation: A Measure for Quantifying Artifacts in Computed Tomography Images With Iterative Reconstructions.

OBJECTIVES: The aims of this study were to introduce the measure noise texture deviation as quantitative parameter for evaluating iterative reconstruction (IR)-specific artifacts in computed tomography (CT) images and to test whether IR-specific artifacts, quantified through this measure, are reduced in advanced modeled IR (ADMIRE) as compared with sinogram-affirmed IR (SAFIRE) images of the liver ex vivo and in patients with hypodense liver lesions. MATERIALS AND METHODS: In the ex vivo study part, an abdominal phantom was used. In the institutional review board-approved in vivo study part, 40 consecutive patients (mean age, 63 years) with hypodense liver lesions undergoing abdominal CT in the portal-venous phase were included. Images were reconstructed with filtered back projection, with the second-generation IR algorithm SAFIRE and with the third-generation IR algorithm ADMIRE. Noise power spectra and noise texture deviation were calculated in the phantom; image noise was measured in the phantom and in patients. Two blinded readers evaluated all image data regarding IR-specific artifacts (plastic-like, blotchy appearance); patient data were evaluated regarding conspicuity and confidence for detecting hypodense liver lesions. RESULTS: Image noise was significantly reduced at increasing IR levels (P < 0.001) with both algorithms, with no significant differences between corresponding strength levels of SAFIRE and ADMIRE (all, P > 0.05). Noise power spectra were similar at corresponding strength levels of SAFIRE and ADMIRE (all, P > 0.05). Noise texture deviation in ADMIRE was reduced compared with corresponding strength levels of SAFIRE (all, P < 0.001) and strongly correlated with subjective IR-specific artifacts (r = 0.88, P < 0.001). Iterative reconstruction-specific artifacts were significantly reduced in ADMIRE compared with that in SAFIRE images at strength levels 3 or greater, both ex vivo and in vivo (all, P < 0.001). There were no significant differences in the readers' ratings of lesion conspicuity and lesion confidence in detecting hypodense liver lesions between SAFIRE and ADMIRE (P > 0.05). Only lesion conspicuity was superior with SAFIRE and ADMIRE compared with filtered back projection (all, P < 0.001). CONCLUSIONS: Noise texture deviation is a quantitative measure reflecting IR-specific artifacts and is reduced in CT images with ADMIRE compared with SAFIRE.

Correlation between Dual-Energy and Perfusion CT in Patients with Hepatocellular Carcinoma.

Purpose To develop a dual-energy contrast media-enhanced computed tomographic (CT) protocol by using time-attenuation curves from previously acquired perfusion CT data and to evaluate prospectively the relationship between iodine enhancement metrics at dual-energy CT and perfusion CT parameters in patients with hepatocellular carcinoma (HCC). Materials and Methods Institutional review board and local ethics committee approval and written informed consent were obtained. The retrospective part of this study included the development of a dual-energy CT contrast-enhanced protocol to evaluate peak arterial enhancement of HCC in the liver on the basis of time-attenuation curves from previously acquired perfusion CT data in 20 patients. The prospective part of the study consisted of an intraindividual comparison of dual-energy CT and perfusion CT data in another 20 consecutive patients with HCC. Iodine density and iodine ratio (iodine attenuation of the lesion divided by iodine attenuation in the aorta) from dual-energy CT and arterial perfusion (AP), portal venous perfusion, and total perfusion (TP) from perfusion CT were compared. Pearson R and linear correlation coefficients were calculated for AP and iodine density, AP and iodine ratio, TP and iodine density, and TP and iodine ratio. Results The dual-energy CT protocol consisted of bolus tracking in the abdominal aorta (threshold, 150 HU; scan delay, 9 seconds). The strongest intraindividual correlations in HCCs were found between iodine density and AP (r = 0.75, P = .0001). Moderate correlations were found between iodine ratio and AP (r = 0.50, P = .023) and between iodine density and TP (r = 0.56, P = .011). No further significant correlations were found. The volume CT dose index (11.4 mGy) and dose-length product (228.0 mGy · cm) of dual-energy CT was lower than those of the arterial phase of perfusion CT (36.1 mGy and 682.3 mGy · cm, respectively). Conclusion A contrast-enhanced dual-energy CT protocol developed by using time-attenuation curves from previously acquired perfusion CT data sets in patients with HCC could show good correlation between iodine density from dual-energy CT with AP from perfusion CT. (©) RSNA, 2016.

Optimizing radiation dose by using advanced modelled iterative reconstruction in high-pitch coronary CT angiography.

PURPOSE: To evaluate the potential of advanced modeled iterative reconstruction (ADMIRE) for optimizing radiation dose of high-pitch coronary CT angiography (CCTA). METHODS: High-pitch 192-slice dual-source CCTA was performed in 25 patients (group 1) according to standard settings (ref. 100 kVp, ref. 270 mAs/rot). Images were reconstructed with filtered back projection (FBP) and ADMIRE (strength levels 1-5). In another 25 patients (group 2), high-pitch CCTA protocol parameters were adapted according to results from group 1 (ref. 160 mAs/rot), and images were reconstructed with ADMIRE level 4. In ten patients of group 1, vessel sharpness using full width at half maximum (FWHM) analysis was determined. Image quality was assessed by two independent, blinded readers. RESULTS: Interobserver agreements for attenuation and noise were excellent (r = 0.88/0.85, p < 0.01). In group 1, ADMIRE level 4 images were most often selected (84%, 21/25) as preferred data set; at this level noise reduction was 40% compared to FBP. Vessel borders showed increasing sharpness (FWHM) at increasing ADMIRE levels (p < 0.05). Image quality in group 2 was similar to that of group 1 at ADMIRE levels 2-3. Radiation dose in group 2 (0.3 ± 0.1 mSv) was significantly lower than in group 1 (0.5 ± 0.3 mSv; p < 0.05). CONCLUSIONS: In a selected population, ADMIRE can be used for optimizing high-pitch CCTA to an effective dose of 0.3 mSv. KEY POINTS: • Advanced modeled IR (ADMIRE) reduces image noise up to 50% as compared to FBP. • Coronary artery vessel borders show an increasing sharpness at higher ADMIRE levels. • High-pitch CCTA with ADMIRE is possible at a radiation dose of 0.3 mSv.

Advanced virtual monoenergetic computed tomography of hyperattenuating and hypoattenuating liver lesions: ex-vivo and patient experience in various body sizes.

OBJECTIVE: To determine the value of advanced virtual monoenergetic images (mono+) from dual-energy computed tomography (CT) of hyperattenuating and hypoattenuating liver lesions in various phantom sizes and patients in comparison with standard monoenergetic images (mono). MATERIALS AND METHODS: Anthropomorphic phantoms simulating 4 patient sizes (S, 300 × 200 mm; M, 350 × 250 mm; L, 400 × 300 mm; and XL, 600 × 450 mm) with a liver insert containing both hyperattenuating and hypoattenuating iodine-containing lesions were imaged with dose-equivalent dual-energy (100/150 Sn kilovolt [peak] [kV{p}]) and single-energy (120 kV[p]) protocols on a 192-slice dual-source CT system. In addition, 4 patients with 3 hypoattenuating and 3 hyperattenuating hepatocellular carcinoma were included and underwent dual-energy CT imaging with the same scanner at similar kV(p) settings (100/150 Sn kV[p]). Images were reconstructed with standard mono and with the mono+ algorithm at 10-kiloelectron volt (keV) intervals from 40 to 190 keV. Attenuation of the liver and lesions were measured, and contrast-to-noise ratios (CNRs) were calculated. Lesion conspicuity was rated by 2 blinded independent readers in all mono and mono+ data sets from 40 to 190 keV using a 5-point Likert scale (1, lowest conspicuity; and 5, highest conspicuity). RESULTS: Attenuation in the liver and in both hyperattenuating and hypoattenuating lesions did not differ between mono and mono+ (P = 0.41-0.49). Noise on mono+ was significantly lower than on mono for all phantom sizes (P < 0.05) and was increasing with phantom size. Hyperattenuating lesion CNR was highest for mono+ images at 40 keV in the S phantom (6.73), with significantly higher CNR for mono+ than for mono and for single energy (120 kV[p]) in all phantom sizes (all P < 0.001) except for the XL phantom. Hypoattenuating lesion CNR was highest for high-keV mono+ being significantly higher than on mono and on single-energy (120 kV[p]) images (all P < 0.001), except for the XL phantom with significantly higher CNR for mono (1.3) compared with mono+ (0.47) and 120 kV(p) (1.26). In patients, CNR curves of hyperattenuating hepatocellular carcinoma were in accordance with the phantom data, whereas hypoattenuating lesions demonstrate varying curves, some being in accordance with findings in phantoms. Interreader agreement for lesion conspicuity was very good (intraclass correlation, 0.95), with higher conspicuity scores for mono+ than for mono and single energy (120 kV[p]) at all phantom sizes (all P < 0.05) and within patients. CONCLUSION: Our ex vivo and patient data demonstrate added value for imaging of both hyperattenuating and hypoattenuating liver lesions with advanced virtual monoenergetic dual-energy CT by decreased noise, increased CNR, and higher lesion conspicuity, although with limitations in XL body sizes.

Prediction model to estimate presence of coronary artery disease: retrospective pooled analysis of existing cohorts.

OBJECTIVES: To develop prediction models that better estimate the pretest probability of coronary artery disease in low prevalence populations. DESIGN: Retrospective pooled analysis of individual patient data. SETTING: 18 hospitals in Europe and the United States. PARTICIPANTS: Patients with stable chest pain without evidence for previous coronary artery disease, if they were referred for computed tomography (CT) based coronary angiography or catheter based coronary angiography (indicated as low and high prevalence settings, respectively). MAIN OUTCOME MEASURES: Obstructive coronary artery disease (≥ 50% diameter stenosis in at least one vessel found on catheter based coronary angiography). Multiple imputation accounted for missing predictors and outcomes, exploiting strong correlation between the two angiography procedures. Predictive models included a basic model (age, sex, symptoms, and setting), clinical model (basic model factors and diabetes, hypertension, dyslipidaemia, and smoking), and extended model (clinical model factors and use of the CT based coronary calcium score). We assessed discrimination (c statistic), calibration, and continuous net reclassification improvement by cross validation for the four largest low prevalence datasets separately and the smaller remaining low prevalence datasets combined. RESULTS: We included 5677 patients (3283 men, 2394 women), of whom 1634 had obstructive coronary artery disease found on catheter based coronary angiography. All potential predictors were significantly associated with the presence of disease in univariable and multivariable analyses. The clinical model improved the prediction, compared with the basic model (cross validated c statistic improvement from 0.77 to 0.79, net reclassification improvement 35%); the coronary calcium score in the extended model was a major predictor (0.79 to 0.88, 102%). Calibration for low prevalence datasets was satisfactory. CONCLUSIONS: Updated prediction models including age, sex, symptoms, and cardiovascular risk factors allow for accurate estimation of the pretest probability of coronary artery disease in low prevalence populations. Addition of coronary calcium scores to the prediction models improves the estimates.

Coronary artery plaques and myocardial ischaemia.

OBJECTIVE: To prospectively examine coronary artery plaques as predictors of myocardial ischaemia using cardiac magnetic resonance (CMR). METHODS: Fifty-two patients (46 men; age 64 ± 10) with suspected coronary artery disease (CAD) referred for catheter coronary angiography (CA) underwent CMR and computed tomography coronary angiography (CTCA). All coronary segments were evaluated for morphological stenosis based on CA. Any plaque according to its composition was assessed based on CTCA. RESULTS: Numbers of total and calcified coronary artery plaques represented the best predictors of myocardial ischaemia (AUC = 0.87; [95%CI: 0.77-0.97] and AUC = 0.87; [95%CI: 0.77-0.96], respectively, p = 0.56) with the total plaque number significantly higher in patients with corresponding ischaemia than those without (p < 0.01, p < 0.05 adjusted for pre-test probability and stenosis). Compared with the AUC of coronary stenosis assessment by CA (AUC = 0.90; [95%CI: 0.80-1.00]), AUCs were equivalent using either the total number or the number of calcified plaques alone (p = 0.73 and p = 0.69). Multivariate logistic regression analyses demonstrated the total plaque number as an independent predictor of ischaemia (odds +20%; [95%CI: 1.096-1.368]), improving a model including clinical probability estimates of CAD (c-statistics, 0.66 to 0.89). CONCLUSION: Coronary artery plaque number according to CTCA is a significant, independent predictor of myocardial ischaemia with similar accuracy to stenosis assessment.

Discrimination between invasive pulmonary aspergillosis and pulmonary lymphoma using CT.

OBJECTIVE: The purpose was to assess the characteristic CT features of invasive pulmonary aspergillosis (IPA) and pulmonary lymphoma (PL) and to analyze the potential to distinguish the two entities using CT. METHODS: The CT images of 70 patients with either proven IPA (n = 35) or PL (n = 35) were evaluated retrospectively and independently by two radiologists (reader 1 [R1] and reader 2 [R2]), analyzing images for presence, number and characteristics of pulmonary nodules and masses, ground-glass opacities, consolidations and other interstitial changes. RESULTS: Interreader agreement was moderate (4/33 CT features), good (9/33) or excellent (20/33). Pulmonary nodules (P = 0.045 [R1], P = 0.001 [R2]), nodules with spiculated outer contours (P < 0.001 [R1], P = 0.001 [R2]), nodules with a halo sign (P < 0.001 [R1 + R2]), nodules with homogeneous (P = 0.030 [R1], P = 0.006 [R2]) and inhomogeneous (P = 0.001 [R1], P < 0.001 [R2]) attenuation patterns, nodules with cavitation (P = 0.006 [R1], P = 0.003 [R2]) and wedge-shaped, pleural-based consolidations (P < 0.001 [R1 + R2]) occurred significantly more often in patients with IPA, while masses without a halo sign (P = 0.03 [R1], P = 0.01 [R2]), lobar consolidations with bronchogram (P = 0.02 [R1 + R2]) and consolidations with homogeneous attenuation patterns (P < 0.001 [R1 + R2]) were found significantly more frequent in PL-patients. CONCLUSIONS: Those CT features can therefore be considered suggestive for either IPA or PL. However, in most cases the diagnosis cannot be made based on CT findings solely because no single feature gained a high sensitivity and specificity concomitantly. Furthermore, the logistic regression did not show a combination that was significantly better than the best univariate predictor.

Dual-energy CT with tin filter technology for the discrimination of renal lesion proxies containing blood, protein, and contrast-agent. An experimental phantom study.

PURPOSE: To differentiate proxy renal cystic lesions containing protein, blood, iodine contrast or saline solutions using dual-energy CT (DECT) equipped with a new tin filter technology (TFT). MATERIALS AND METHODS: 70 proxies (saline, protein, blood and contrast agent) were placed in unenhanced and contrast-enhanced kidney phantoms. DECT was performed at 80/140 kV with and without tin filtering. Two readers measured the CT attenuation values in all proxies twice. An 80/140 kV ratio was calculated. RESULTS: All intra- and interobserver agreements were excellent (r = 0.93-0.97; p < 0.001). All CT attenuation values were significantly higher in the enhanced than in the unenhanced setting (p < 0.05; average increase, 12.5 ± 3.6 HU), while the ratios remained similar (each, p > 0.05). The CT attenuation of protein, blood and contrast agent solution differed significantly with tin filtering (p < 0.01-0.05). Significant differences were found between the ratios of protein and blood compared to contrast medium solution (each, p < 0.05) and between the ratios of protein and blood in both phantoms with tin filtering (each, p < 0.05). CONCLUSION: DECT allows discrimination between a proxy renal lesion containing contrast agent and lesions containing protein and blood through their different attenuation at 80 kV and 140 kV. Further discrimination between protein and blood containing proxies is possible when using a tin filter.

High-pitch 128-slice dual-source CT for the assessment of coronary stents in a phantom model.

RATIONALE AND OBJECTIVES: To evaluate the quality of stent lumen delineation using dual-source computed tomography (DSCT) in the standard-pitch mode (SP) as compared to the high-pitch mode (HP) in a phantom study. MATERIALS AND METHODS: Forty different coronary stents placed in plastic tubes filled with contrast agent were imaged with a second generation DSCT system in a SP (pitch 0.23) and HP (pitch 3.4) mode in orientations of 0°, 45°, and 90° relative to the z-axis. Two observers independently measured the in-stent lumen and the attenuation values in the center of the stents. The artificial lumen narrowing (ALN) was calculated using the measured in-stent lumen and the nominal diameter of the plastic tube. RESULTS: Interobserver correlation was excellent for in-stent lumen (0.86) and attenuation measurements (0.91). There was no significant difference neither for ALN (SP: 54.7-62.8%; HP: 55.8-64.0%) nor attenuation (SP: 356-395 Hounsfield units [HU]; HP: 352-384 HU) between SP and HP mode. For both modes, the orientation of the stent relative to the z-axis significantly affected ALN and attenuation (each P < .001). CT volume dose index was significantly lower using HP mode as compared to SP mode (P < .001). CONCLUSION: The HP mode in DSCT provides visualization of the coronary in-stent lumen comparable to that measured in SP mode while reducing applied radiation dose in a stationary phantom model.

Low-dose, 128-slice, dual-source CT coronary angiography: accuracy and radiation dose of the high-pitch and the step-and-shoot mode.

OBJECTIVE: To compare the diagnostic accuracy and radiation doses of two low-dose protocols for coronary artery imaging with second-generation, dual-source CT in comparison with catheter angiography (CA). DESIGN, SETTING AND PATIENTS: Prospective, single-centre study conducted in a referral centre enrolling 100 patients with low-to-intermediate risk and suspicion of coronary artery disease. All patients underwent contrast-enhanced, 128-slice, dual-source CT coronary angiography and CA. Patients were randomly assigned to two different low-dose CT protocols (each 100 kV/320 mA): in group A (n=50), CT was performed using the prospectively electrocardiography (ECG)-gated step-and-shoot (SAS) mode; in group B (n=50), CT was performed using the prospectively ECG-gated high-pitch mode (pitch 3.4). The image quality and presence or absence of significant coronary stenosis in all coronary segments were evaluated by two blinded and independent observers. CA served as the standard of reference. RESULTS: Sixty-one significant stenoses were found in group A, and 69 in group B. There was no significant difference in age (group A, 62+/-8 yrs; group B, 63+/-8 yrs; p=0.72), body mass index (group A, 26.4+/-3.1 kg/m2; group B, 25.9+/-2.8 kg/m2; p=0.41) and heart rate (HR) (group A, 58+/-8 bpm; group B, 56+/-10 bpm; p=0.66) between the groups. Diagnostic image quality was obtained in 98.6% (651/660) of segments in group A and in 98.9% (642/649) in group B, with no significant differences between groups. Sensitivity, specificity and positive and negative predictive values were 94%, 91%, 85% and 97% per-patient in group A, and 93%, 94%, 89% and 97% per-patient in group B (no significant differences). The effective radiation dose in group B (0.9+/-0.1 mSv) was significantly (p<0.01) lower than that in group A (1.4+/-0.4 mSv). CONCLUSIONS: Both the high-pitch and the SAS mode for low-dose CT coronary angiography provide high accuracy for the assessment of significant coronary stenoses, while the high-pitch mode further significantly lowers the radiation dose.

Quantitative computed tomography liver perfusion imaging using dynamic spiral scanning with variable pitch: feasibility and initial results in patients with cancer metastases.

PURPOSE: To assess the feasibility and image quality of computed tomography (CT) liver perfusion imaging using an adaptive 4D spiral-mode, developed to extend the z-axis coverage, and to report initial qualitative and quantitative results in patients with cancer metastases. MATERIALS AND METHODS: A total of 21 patients with liver metastases of various origins underwent CT perfusion imaging (100 kV and 150 mAs/rot) using a 4D spiral-mode with single-source 64-slice CT (n = 7) with a scan range of 6.7 cm (protocol A: 16 cycles, 46.5 seconds examination time), or dual-source 128-slice CT with a scan range of 14.8 cm (protocol B: 16 cycles, 46.5 seconds examination time, n = 7; protocol C: 12 cycles, 51.0 seconds examination time, n = 7). Ability to suspend respiration during perfusion imaging was monitored. Two independent readers assessed image quality on a 4-point scale, both before and after motion correction, and performed a qualitative (ie, arterial enhancement pattern and enhancement change over time) and quantitative perfusion (ie, arterial liver perfusion [ALP]; portal-venous perfusion [PVP]; hepatic perfusion index [HPI]) analysis. RESULTS: Of 21 patients, 7 (33%) could suspend respiration throughout the perfusion study and 14 (67%) resumed shallow breathing during the perfusion scan. The 21 patients had a total of 88 metastases. The scan range of protocol A covered at least 1 metastasis in all patients (total 20/34 [58.8%] metastases). The scan range of protocol B and C covered 53 of 54 (98.1%) metastases, whereas one metastasis in segment VIII was only partially imaged. Image quality was diagnostic both before and after motion correction, whereas being significantly better after motion correction (P < 0.001). Qualitative perfusion analysis of 67 metastases revealed diffuse arterial enhancement in 3 (4.5%), sparse enhancement in 11 (16.4%), peripheral-nodular enhancement in 9 (13.4%), rim-like enhancement in 15 (22.4%), and none in 29 (43.3%) metastases. Enhancement over time of 67 metastases showed a centripetal progression in 6 (8.9%), sustained portal phase in 16 (23.9%), wash-out in 16 (23.9%), and none in 29 (43.3%) metastases. Quantitative perfusion analysis revealed significantly higher arterial liver perfusion and HPI in metastases and metastasis borders than in adjacent normal liver tissue (P < 0.001 each). Portal-venous perfusion was significantly lower in metastases and metastasis borders than in normal liver tissue (P < 0.001). There were no significant differences in image quality and qualitative perfusion analysis between the 3 protocols (P = n.s.). Calculated effective radiation doses were 13.4 mSv for protocol A, 30.7 mSv for protocol B, and 23.0 mSv for protocol C. CONCLUSION: CT perfusion imaging of the liver using the 4D spiral-mode is feasible with diagnostic image quality, and enables the reliable qualitative and quantitative analysis of the normal and metastatic liver parenchyma. Radiation dose issues must be considered when determining the scan range, number of cycles, and scan duration of the perfusion CT protocol.

Low dose high-pitch spiral acquisition 128-slice dual-source computed tomography for the evaluation of coronary artery bypass graft patency.

OBJECTIVES: To prospectively assess electrocardiography (ECG)-synchronized dual-source computed tomography (CT) in high-pitch spiral acquisition mode for the evaluation of coronary artery bypass graft (CABG) patency regarding image quality and radiation dose. MATERIALS AND METHODS: Fifty consecutive patients (47 men, age 69.6 +/- 9.6 years, body mass index 26.6 +/- 3.5 kg/m) underwent clinically indicated 128-slice dual-source CT angiography of the entire thorax for the evaluation of graft patency after CABG surgery using a prospectively ECG-synchronized high-pitch spiral acquisition mode (pitch 3.2; 100 kV tube voltage, 0.28 seconds gantry rotation time). Heart rates (HR) were noted. Image quality of the proximal anastomosis, graft body, distal anastomosis, and postanastomotic coronary artery were graded by 2 independent readers on a 3-point scale (1 = excellent, 2 = moderate, and 3 = poor/nondiagnostic). Quantitative image quality parameters (noise, contrast-to-noise ratio) were measured. Effective radiation doses were calculated. RESULTS: Mean HR was 76 +/- 19 bpm (range, 45-135 bpm). Median scan time was 0.9 seconds (range, 0.76-1.17 seconds), mean scan length was 349 +/- 38 mm (range, 294-452 mm). A total of 125 CABG (54% arterial grafts) and 465 vessel sections were analyzed. Twelve grafts were proximally occluded. Diagnostic image quality was obtained in 462 (99.4%) sections. Of those, image quality was excellent in 397 (85.4%) and moderate in 65 (14.0%). All 3 (0.6%) nondiagnostic sections affected the distal anastomosis due to motion artifacts caused by high HR. Separating the study population by the median HR of 71 bpm, image quality of the distal anastomosis regarding motion artifacts was significantly lower at high HR (P < 0.05). Interobserver agreement in grading image quality of graft segments was good (kappa = 0.77). Image noise in the aorta was 36.9 +/- 8.1, contrast-to-noise ratio was 13.1 +/- 4.2. Effective radiation dose was 2.3 +/- 0.3 mSv. CONCLUSIONS: The patency of coronary artery bypass grafts can be assessed with decreasing image quality at high HR in high-pitch prospectively ECG-synchronized thoracic 128-slice dual-source CT angiography at a low radiation dose.

Performance of dual-energy CT with tin filter technology for the discrimination of renal cysts and enhancing masses.

RATIONALE AND OBJECTIVES: To assess the performance of dual-energy computed tomography (DECT) equipped with the new tin filter technology to classify phantom renal lesions as cysts or enhancing masses. MATERIALS AND METHODS: Forty spherical lesion proxies ranging in diameter from 6 to 27 mm were filled with either distilled water (n = 10) representing cysts or titrated iodinated contrast solutions with a concentration of 0.45 (n = 10), 0.9 (n = 10), and 1.8 mg/mL (n = 10) representing enhancing masses. The lesion proxies were placed in a 12-cm diameter renal phantom containing minced beef and submerged in a 28-cm water bath. DECT was performed using the new dual-source CT system (Definition Flash, Siemens Healthcare, Forchheim, Germany) allowing for an improved energy separation by using a tin filter. DECT was performed at tube voltages of 140/80 kV without the tin filter (protocol A) and with tin filter (protocol B). The tube current time product was selected in each protocol to achieve a constant CTDI (computed tomography dose index) with both protocols of 19 mGy (full dose), 9.5 mGy (half dose), and 4.8 mGy (quarter dose). Two blinded readers classified each lesion as a cyst or enhancing mass by using iodine overlay (IO) images. One reader measured the CT numbers of each lesion at 120 kV, in the IO, linear blending (LB), and virtual noncontrast (VNC) images. RESULTS: The CT numbers of the lesions at 120 kV were 0.1 +/- 0.7 HU (0 mg iodine/mL), 9.1 +/- 0.7 HU (0.45 mg/mL), 18.1 +/- 1.4 HU (0.9 mg/mL), and 37.6 +/- 1.6 HU (1.8 mg/mL). Mean diameter of the lesion proxies filled with water or different iodine concentrations was similar (P = 0.38). Image noise was not significantly different in protocols A and B at the corresponding dose levels. At full dose, protocol A had a sensitivity of 93% and a specificity of 60% for discriminating renal lesions. Sensitivity and specificity declined to 84% and 38% at quarter dose. In protocol B, sensitivity was 100% and specificity was 90% at full dose and 93% and 70% at quarter dose. All misclassifications occurred in cyst or low iodine concentration (0.45 mg/mL) lesion proxies. The differences between CT numbers at 120 kV and in the IO, VNC, and AW (average weighted) images were significantly lower in protocol B compared to protocol A (each P < .05). CONCLUSIONS: DECT using the tin filter results in an improved sensitivity and specificity for discriminating renal cysts from enhancing masses in a kidney phantom model and demonstrates higher dose efficiency as compared to former dual energy technology without tin filters.

Low-dose CT and cardiac MR for the diagnosis of coronary artery disease: accuracy of single and combined approaches.

To prospectively compare the diagnostic performance of low-dose computed tomography coronary angiography (CTCA) and cardiac magnetic resonance imaging (CMR) and combinations thereof for the diagnosis of significant coronary stenoses. Forty-three consecutive patients with known or suspected coronary artery disease underwent catheter coronary angiography (CA), dual-source CTCA with prospective electrocardiography-gating, and cardiac CMR (1.5 Tesla). The following tests were analyzed: (1) low-dose CTCA, (2) adenosine stress-rest perfusion-CMR, (3) late gadolinium enhancement (LGE), (4) perfusion-CMR and LGE, (5) low-dose CTCA combined with perfusion-CMR, (5) low-dose CTCA combined with late gadolinium-enhancement, (6) low-dose CTCA combined with perfusion-CMR and LGE. CA served as the standard of reference. CA revealed >50% diameter stenoses in 68/129 (57.7%) coronary arteries in 29/43 (70%) patients. In the patient-based analysis, sensitivity, specificity, NPV and PPV of low-dose CTCA for the detection of significant stenoses were 100, 92.9, 100 and 96.7%, respectively. For perfusion-CMR and LGE, sensitivity, specificity, NPV, PPV, and accuracy were 89.7, 100, 82.4, and 100%, respectively. In the artery-based analysis, sensitivity and NPV of low-dose CTCA was significantly (P < 0.05) higher than that of perfusion-CMR and LGE. All combinations of low-dose CTCA and perfusion-CMR and/or LGE did not improve the diagnostic performance when compared to low-dose CTCA alone. Taking CA as standard of reference, low-dose CTCA outperforms CMR with regard to sensitivity and NPV, whereas CMR is more specific and has a higher PPV than low-dose CTCA.

Left atrial appendage clip occlusion: early clinical results.

OBJECTIVE: Atrial fibrillation puts patients at significant risk for embolic stroke originating from the left atrial appendage. Few means are available for safe, effective, and durable left atrial appendage occlusion. A new clip device was evaluated with regard to safety and effectiveness for epicardial left atrial appendage occlusion. METHODS: Patients with atrial fibrillation undergoing elective cardiac surgery through a median sternotomy were enrolled for concomitant epicardial clip placement. Early postoperative and 3-month follow-up computed tomography studies were used to assess clip stability and left atrial appendage perfusion. RESULTS: From September 2007 to December 2008, 34 patients underwent successful clip placement. No device-related complications occurred. Operative mortality was 8.8% and not study or device related. Deployment was rapid, and left atrial appendage occlusion was confirmed by intraoperative transesophageal echocardiography in all patients. In addition to excellent clinical outcomes (no stroke/transient ischemic attack), serial computed tomography demonstrated stable clip location and appendage perfusion at 3 months in all patients. CONCLUSION: Safe, effective, and durable left atrial appendage occlusion can easily be achieved with this new clip. Further trials are necessary to evaluate the role of the left atrial appendage occlusion in stroke prevention.

Guided review by frequent itemset mining: additional evidence for plaque detection.

PURPOSE: A guided review process to support manual coronary plaque detection in computed tomography coronary angiography (CTCA) data sets is proposed. The method learns the spatial plaque distribution patterns by using the frequent itemset mining algorithm and uses this knowledge to predict potentially missed plaques during detection. MATERIALS AND METHODS: Plaque distribution patterns from 252 manually labeled patients who underwent CTCA were included. For various cross-validations a labeling with missing plaques was created from the initial manual ground truth labeling. Frequent itemset mining was used to learn the spatial plaque distribution patterns in form of association rules from a training set. These rules were then applied on a testing set to search for segments in the coronary tree showing evidence of containing unlabeled plaques. The segments with potentially missed plaques were finally reviewed for the existence of plaques. The proposed guided review was compared to a weighted random approach that considered only the probability of occurrence for a plaque in a specific segment and not its spatial correlation to other plaques. RESULTS: Guided review by frequent itemset mining performed significantly better (p < 0.001) than the reference weighted random approach in predicting coronary segments with initially missed plaques. Up to 47% of the initially removed plaques were refound by only reviewing 4.4% of all possible segments. CONCLUSIONS: The spatial distribution patterns of atherosclerosis in coronary arteries can be used to predict potentially missed plaques by a guided review with frequent itemset mining. It shows potential to reduce the intra- and inter-observer variability.

The heart of patients with aortic aneurysms: evidence from cardiac computed tomography.

To determine in patients with abdominal aortic aneurysm (AAA) the coronary calcium burden and prevalence of coronary artery disease (CAD) in relation to cardiovascular risk factors, and to assess the left ventricular (LV) function using cardiac computed tomography (CT). Sixty consecutive patients (six females; 72.2+/-9.0 years) with AAA underwent dual-source CT calcium scoring and coronary angiography prior to AAA repair. In the 60 patients, the Framingham risk score (FRS) ranged from 5-43%. Twenty patients (33%) were at low, 16 (27%) at intermediate, and 24 (40%) at high risk for cardiovascular disease. The median Agatston score (AS) was 393 (0-3538). No significant correlation was found between AS and FRS (P=0.76). 846/851 coronary segments (99%) in 57/60 patients (95%) were depicted with a diagnostic image quality. Significant stenoses were found in 132/846 segments (16%) in 33/60 patients (55%). Five patients (8%) with significant coronary artery stenosis showed reduced LV function [ejection fraction (EF)<50%]. The extent of CAD was significantly correlated with AS (r=0.43, P<0.01), whereas no correlation was found for FRS (P=0.55). Cardiac CT is feasible in patients with AAA and allows for the assessment of coronary calcium, coronary stenoses, and LV function. The calcium burden and coronary stenoses assessment with cardiac CT provides incremental information beyond traditional cardiovascular risk factors alone.

Accuracy of dual-source computed tomography coronary angiography: evaluation with a standardised protocol for cardiac surgeons.

BACKGROUND: This study assesses the accuracy of the new dual-source computed tomography (DSCT) for detection of coronary artery disease (CAD) compared with invasive coronary angiography (ICA) with a specifically designed data presentation protocol for cardiac surgeons. METHODS: Forty patients (30 males/10 females) underwent ICA and DSCT. Best-quality images were prepared by radiologists. Evaluation of 12 segments of significant coronary stenosis was done by two cardiac surgeons with a data presentation protocol including different coronary views in two-/three-dimensional (2D/3D) images. No beta-blockers were administered prior to DSCT. RESULTS: ICA revealed CAD in 21 patients and valvular disease but no CAD in 19 patients. In DSCT, 20/21 patients were diagnosed with CAD (at least one significant stenosis per patient). In 11/21 patients, all 12 segments were assessed correctly; in 7/21 patients one segment and in 3/21 patients two segments were evaluated incorrectly. Of all 21 patients with CAD, 239/252 segments (95%) were correctly evaluated. In 18/19 patients without CAD, DSCT correctly ruled-out the ICA results in 226/228 segments (99%). In total, 465/480 segments were correctly assessed (97%). Of 480 segments, only six were considered not assessable. DSCT assessments of the segments showed a sensitivity of 91%, specificity of 99%, a positive predictive value of 92% and a negative predictive value of 99%. CONCLUSIONS: The accuracy of DSCT coronary angiography especially for exclusion of CAD is promising. The introduced data presentation protocol allows for the independent evaluation by cardiac surgeons after pre-arrangement from the radiologists.