[Use of contrast-enhanced mammography for diagnosis of breast cancer]. / Einsatz der kontrastmittelverstärkten Mammographie in der Brustkrebsdiagnostik.

BACKGROUND: Contrast-enhanced mammography (CEM) is an imaging method that is able to improve visualization of intramammary tumors after peripheral venous administration of an iodine-containing contrast medium (ICM). OBJECTIVES AND METHODS: The current significance of CEM is discussed. RESULTS: Studies were able to show an advantage of CEM in the diagnosis of breast cancer compared to mammography, especially for women with dense breasts. Indications for CEM currently depend on the availability of magnetic resonance imaging (MRI). If MRI is available, CEM is indicated in those cases when MRI cannot be performed. Use of CEM for breast cancer screening is currently viewed critically. This view can change when results and updated assessments of large CEM studies in Europe and USA become available. Patients must be informed about the use of an ICM. As ICM administration for CEM is carried out in a similar manner to established imaging methods, the authors expect the use of ICM for CEM to be unproblematic as long as general contraindications are adhered to. CONCLUSIONS: In the future, CEM could have greater importance for the diagnosis of breast cancer, as this imaging method has diagnostic advantages compared to conventional mammography. A great advantage of CEM is its availability. For those who use breast MRI, CEM is helpful when MRI is not feasible due to contraindications or other reasons.

Automation Bias in Mammography: The Impact of Artificial Intelligence BI-RADS Suggestions on Reader Performance.

Background Automation bias (the propensity for humans to favor suggestions from automated decision-making systems) is a known source of error in human-machine interactions, but its implications regarding artificial intelligence (AI)-aided mammography reading are unknown. Purpose To determine how automation bias can affect inexperienced, moderately experienced, and very experienced radiologists when reading mammograms with the aid of an artificial intelligence (AI) system. Materials and Methods In this prospective experiment, 27 radiologists read 50 mammograms and provided their Breast Imaging Reporting and Data System (BI-RADS) assessment assisted by a purported AI system. Mammograms were obtained between January 2017 and December 2019 and were presented in two randomized sets. The first was a training set of 10 mammograms, with the correct BI-RADS category suggested by the AI system. The second was a set of 40 mammograms in which an incorrect BI-RADS category was suggested for 12 mammograms. Reader performance, degree of bias in BI-RADS scoring, perceived accuracy of the AI system, and reader confidence in their own BI-RADS ratings were assessed using analysis of variance (ANOVA) and repeated-measures ANOVA followed by post hoc tests and Kruskal-Wallis tests followed by the Dunn post hoc test. Results The percentage of correctly rated mammograms by inexperienced (mean, 79.7% ± 11.7 [SD] vs 19.8% ± 14.0; P < .001; r = 0.93), moderately experienced (mean, 81.3% ± 10.1 vs 24.8% ± 11.6; P < .001; r = 0.96), and very experienced (mean, 82.3% ± 4.2 vs 45.5% ± 9.1; P = .003; r = 0.97) radiologists was significantly impacted by the correctness of the AI prediction of BI-RADS category. Inexperienced radiologists were significantly more likely to follow the suggestions of the purported AI when it incorrectly suggested a higher BI-RADS category than the actual ground truth compared with both moderately (mean degree of bias, 4.0 ± 1.8 vs 2.4 ± 1.5; P = .044; r = 0.46) and very (mean degree of bias, 4.0 ± 1.8 vs 1.2 ± 0.8; P = .009; r = 0.65) experienced readers. Conclusion The results show that inexperienced, moderately experienced, and very experienced radiologists reading mammograms are prone to automation bias when being supported by an AI-based system. This and other effects of human and machine interaction must be considered to ensure safe deployment and accurate diagnostic performance when combining human readers and AI. © RSNA, 2023 Supplemental material is available for this article. See also the editorial by Baltzer in this issue.

Correlation of CT-data derived from multiparametric dual-layer CT-maps with immunohistochemical biomarkers in invasive breast carcinomas.

OBJECTIVE: To examine the correlation of quantitative measurements from material decomposition maps calculated from dual-layer CT (DLCT)-image datasets with immunohistochemical biomarkers of invasive breast carcinomas. MATERIAL AND METHODS: All patients at the University Breast Cancer Center who underwent a clinically indicated dual-layer CT-scan for staging of invasive ductal breast carcinoma from 01/2016 to 07/2020 were prospectively included. Iodine concentration maps and maps of the effective atomic numbers (Zeffective) were reconstructed from the image datasets. ROI-based evaluations of the index tumors and predefined references tissues for normalization were performed semi-automatically in identical anatomical positions using dedicated evaluation software. Statistical analysis was essentially descriptive using Spearmans rank correlation and (multivariable) partial correlation. RESULTS: Bivariate showed statistically significant correlations of iodine contents (r = -0.154/-0.202/0.180, p = 0.039/0.006/0.015), and Zeffective-values (r = -0.158/-0.199/0.179, p = 0.034/0.007/0.016) for all 184 carcinomas and the subgroup of 168 invasive ductal carcinomas. The results were confirmed by multivariate analyses with "age", "diameter" and "ACR-grade" as possible confounders. Normalization of the measured target values with those in the aorta confirmed significant correlations of iodine content and Zeffective compared to Estrogen (r = 0.174, p = 0.019), Progesteron (r = 0.168/0.177, p = 0.024/0.017), and HER2 receptor expression (r = -0.222/-0.184, p = 0.003/0.013). All CT-parameters showed significant correlations with immunohistochemical subtyping (r = 0.191/0.192, p = 0.010). CONCLUSIONS: Our preliminary results indicate that iodine content and Zeffective-values derived from DLCT-examinations correlate with hormone receptor expression in invasive breast carcinomas. Assignments to benign entities already seam feasible in clinical routine CT-diagnostics. After further investigations iodine content and Zeffective may be translated as diagnostical and prognostical biomarkers into clinical routine in the long term.

Long-term survival of a BRCA2 mutation carrier following second ovarian cancer relapse using PARPi therapy: A case report.

BRCA1/2 mutation carriers have lifelong increased risks of developing breast and ovarian cancer. Due to the lack of efficient ovarian cancer screening, patients mainly present when the tumors are at an advanced stage, and the long-term survival is poor. The application of poly(ADP-ribose) polymerase (PARP) inhibitors (PARPi) has been used in personalized cancer treatment. Specific strategies to improve the outcome of patients are available and mainly include targeting of BRCA1 or BRCA2 mutations. The aim of the present study was to report the case of a 67-year old BRCA2 mutation carrier, who was enrolled in 2010 in one of the first PARPi studies (ICEBERG2). The patient exhibited second ovarian cancer (OC) relapse following a platinum-free interval of 5 months. The third-line treatment with olaparib monotherapy was initiated in January 2011, without prior surgery or chemotherapy, and achieved a persistent response. The patient demonstrated an unprecedented long-term survival of >9 years under PARPi monotherapy after the second relapse of OC. The data of the present case report support the use of PARPi as a well-tolerated and effective long-term treatment, even for patients with unfavourable prognostic characteristics, such as platinum resistance, without immediately preceding optimal cytoreduction. However, further studies are required to provide more insight into the selection of patients for favourable maintenance treatment.

Consensus Recommendations of the German Consortium for Hereditary Breast and Ovarian Cancer.

Background: The German Consortium for Hereditary Breast and Ovarian Cancer (GC-HBOC) has established a multigene panel (TruRisk®) for the analysis of risk genes for familial breast and ovarian cancer. Summary: An interdisciplinary team of experts from the GC-HBOC has evaluated the available data on risk modification in the presence of pathogenic mutations in these genes based on a structured literature search and through a formal consensus process. Key Messages: The goal of this work is to better assess individual disease risk and, on this basis, to derive clinical recommendations for patient counseling and care at the centers of the GC-HBOC from the initial consultation prior to genetic testing to the use of individual risk-adapted preventive/therapeutic measures.

Automated mapping and N-Staging of thoracic lymph nodes in contrast-enhanced CT scans of the chest using a fully convolutional neural network.

PURPOSE: To develop a deep-learning (DL)-based approach for thoracic lymph node (LN) mapping based on their anatomical location. METHOD: The training-and validation-dataset included 89 contrast-enhanced computed tomography (CT) scans of the chest. 4201 LNs were semi-automatically segmented and then assigned to LN levels according to their anatomical location. The LN level classification task was addressed by a multi-class segmentation procedure using a fully convolutional neural network. Mapping was performed by firstly determining potential level affiliation for each voxel and then performing majority voting over all voxels belonging to each LN. Mean classification accuracies on the validation data were calculated separately for each level and overall Top-1, Top-2 and Top-3 scores were determined, where a Top-X score describes how often the annotated class was within the top-X predictions. To demonstrate the clinical applicability of our model, we tested its N-staging capabilities in a simulated clinical use case scenario assuming a patient diseased with lung cancer. RESULTS: The artificial intelligence(AI)-based assignment revealed mean classification accuracies of 86.36 % (Top-1), 94.48 % (Top-2) and 96.10 % (Top-3). Best accuracies were achieved for LNs in the subcarinal level 7 (98.31 %) and axillary region (98.74 %). The highest misclassification rates were observed among LNs in adjacent levels. The proof-of-principle application in a simulated clinical use case scenario for automated tumor N-staging showed a mean classification accuracy of up to 96.14 % (Top-1). CONCLUSIONS: The proposed AI approach for automatic classification of LN levels in chest CT as well as the proof-of-principle-experiment for automatic N-staging, revealed promising results, warranting large-scale validation for clinical application.

Automated detection and segmentation of thoracic lymph nodes from CT using 3D foveal fully convolutional neural networks.

BACKGROUND: In oncology, the correct determination of nodal metastatic disease is essential for patient management, as patient treatment and prognosis are closely linked to the stage of the disease. The aim of the study was to develop a tool for automatic 3D detection and segmentation of lymph nodes (LNs) in computed tomography (CT) scans of the thorax using a fully convolutional neural network based on 3D foveal patches. METHODS: The training dataset was collected from the Computed Tomography Lymph Nodes Collection of the Cancer Imaging Archive, containing 89 contrast-enhanced CT scans of the thorax. A total number of 4275 LNs was segmented semi-automatically by a radiologist, assessing the entire 3D volume of the LNs. Using this data, a fully convolutional neuronal network based on 3D foveal patches was trained with fourfold cross-validation. Testing was performed on an unseen dataset containing 15 contrast-enhanced CT scans of patients who were referred upon suspicion or for staging of bronchial carcinoma. RESULTS: The algorithm achieved a good overall performance with a total detection rate of 76.9% for enlarged LNs during fourfold cross-validation in the training dataset with 10.3 false-positives per volume and of 69.9% in the unseen testing dataset. In the training dataset a better detection rate was observed for enlarged LNs compared to smaller LNs, the detection rate for LNs with a short-axis diameter (SAD) ≥ 20 mm and SAD 5-10 mm being 91.6% and 62.2% (p < 0.001), respectively. Best detection rates were obtained for LNs located in Level 4R (83.6%) and Level 7 (80.4%). CONCLUSIONS: The proposed 3D deep learning approach achieves an overall good performance in the automatic detection and segmentation of thoracic LNs and shows reasonable generalizability, yielding the potential to facilitate detection during routine clinical work and to enable radiomics research without observer-bias.

Plasma interferon-γ-inducible protein 10 (IP-10) levels correlate with disease severity and paradoxical reactions in extrapulmonary tuberculosis.

BACKGROUND: With 1.5 million deaths worldwide in 2018, tuberculosis (TB) remains a major global public health problem. While pulmonary TB (PTB) is the most common manifestation, the proportion of extrapulmonary TB (EPTB) is increasing in low-burden countries. EPTB is a heterogeneous disease entity posing diagnostic and management challenges due to the lack of reliable biomarkers. In this study, we prospectively evaluated clinical data and treatment response which were correlated with different biomarkers. METHODS: The study was conducted at the University Hospital of Cologne. 20 patients with EPTB were enrolled. We analyzed plasma interferon-γ-inducible protein 10 (IP-10) levels in plasma by ELISA for up to 12 months of treatment. In addition, the QuantiFERON®-TB Gold Plus (QFT® Plus) test was performed during the course of treatment. Clinical data were assessed prospectively and correlated with QFT® Plus and IP-10 levels. RESULTS: Plasma IP-10 levels were found to be significantly increased (p < 0.001) in patients with extensive disease compared to patients with limited disease (cervical lymph node TB) or healthy controls. In patients with clinically confirmed paradoxical reaction (PR), a further increase of IP-10 was noted. IFN-γ measured by the QFT® Plus test did not decrease significantly during the course of treatment. Of note, in four EPTB patients (20%) without radiographic pulmonary involvement, sputum culture was positive for Mycobacterium tuberculosis. CONCLUSION: Our data demonstrate that IP-10 may be a valuable biomarker for estimation of disease severity in EPTB and monitoring of the disease course in extensive forms. However, IP-10 may be less suitable for diagnosis and monitoring of EPTB patients with limited disease. The QFT® Plus test does not appear to be a suitable marker for therapy monitoring. Sputum should be examined in EPTB patients even in case of normal diagnostic imaging of the chest.

Quantitative accuracy of virtual non-contrast images derived from spectral detector computed tomography: an abdominal phantom study.

Dual-energy CT allows for the reconstruction of virtual non-contrast (VNC) images. VNC images have the potential to replace true non-contrast scans in various clinical applications. This study investigated the quantitative accuracy of VNC attenuation images considering different parameters for acquisition and reconstruction. An abdomen phantom with 7 different tissue types (different combinations of 3 base materials and 5 iodine concentrations) was scanned using a spectral detector CT (SDCT). Different phantom sizes (S, M, L), volume computed tomography dose indices (CTDIvol 10, 15, 20 mGy), kernel settings (soft, standard, sharp), and denoising levels (low, medium, high) were tested. Conventional and VNC images were reconstructed and analyzed based on regions of interest (ROI). Mean and standard deviation were recorded and differences in attenuation between corresponding base materials and VNC was calculated (VNCerror). Statistic analysis included ANOVA, Wilcoxon test and multivariate regression analysis. Overall, the VNCerror was - 1.4 ± 6.1 HU. While radiation dose, kernel setting, and denoising level did not influence VNCerror significantly, phantom size, iodine content and base material had a significant effect (e.g. S vs. M: - 1.2 ± 4.9 HU vs. - 2.1 ± 6.0 HU; 0.0 mg/ml vs. 5.0 mg/ml: - 4.0 ± 3.5 HU vs. 5.1 ± 5.0 HU and 35-HU-base vs. 54-HU-base: - 3.5 ± 4.4 HU vs. 0.7 ± 6.5; all p ≤ 0.05). The overall accuracy of VNC images from SDCT is high and independent from dose, kernel, and denoising settings; however, shows a dependency on patient size, base material, and iodine content; particularly the latter results in small, yet, noticeable differences in VNC attenuation.

Semiautomated Renal Cortex Volumetry in Spectral Computed Tomography: Effect of Monoenergetic Reconstructions on Measurement Precision and Interobserver Variability.

OBJECTIVE: The aim of this study was to determine the influence of virtual monoenergetic images (vMEIs) on renal cortex volumetry (RCV) and estimation of split-renal function. METHODS: Twenty-five patients (mean ± SD, 64.7 ± 9.9 years) underwent a contrast-enhanced dual-layer spectral detector computed tomography. Images were reconstructed with a reference standard (iterative model reconstruction, IMRRef), a newly spectral detector computed tomography algorithm (SPcon) and vMEI at 40, 60, 80, 100, and 120 keV. Two blinded independent readers performed RCV on all data sets with a semiautomated tool. RESULTS: Total kidney volume was up to 15% higher in vMEI at 40/60 keV compared with IMRRef (P < 0.001). Total kidney volume with vMEI at 80/100 keV was similar to IMRRef (P < 0.001). Split-renal function was similar in all reconstructions at approximately 50% ± 3%. Bland-Altman analysis showed no significant differences (P > 0.05), except for 40 keV versus SPcon (P < 0.05). The time required to perform RCV was reasonable, approximately 4 minutes, and showed no significant differences among reconstructions. Interreader agreement was greatest with vMEI at 80 keV (r = 0.68; 95% confidence interval, 0.39-0.85; P < 0.0002) followed by IMRRef images (r = 0.67; 95% confidence interval, 0.37-0.84; P < 0.0003). IMRRef showed the highest mean Hounsfield unit for cortex/medulla of 223.4 ± 73.7/62.5 ± 19.7 and a ratio of 3.7. CONCLUSIONS: Semiautomated RCV performed with vMEI and IMRRef/SPcon is feasible and showed no clinically relevant differences with regard to split-renal function. Low-kiloelectron volt vMEI showed greater tissue contrast and total kidney volume but no benefit for RCV. Moderate-kiloelectron volt vMEI (80 keV) results were similar to IMRRef with a faster postprocessing time.

Overcoming EGFRG724S-mediated osimertinib resistance through unique binding characteristics of second-generation EGFR inhibitors.

The emergence of acquired resistance against targeted drugs remains a major clinical challenge in lung adenocarcinoma patients. In a subgroup of these patients we identified an association between selection of EGFRT790M-negative but EGFRG724S-positive subclones and osimertinib resistance. We demonstrate that EGFRG724S limits the activity of third-generation EGFR inhibitors both in vitro and in vivo. Structural analyses and computational modeling indicate that EGFRG724S mutations may induce a conformation of the glycine-rich loop, which is incompatible with the binding of third-generation TKIs. Systematic inhibitor screening and in-depth kinetic profiling validate these findings and show that second-generation EGFR inhibitors retain kinase affinity and overcome EGFRG724S-mediated resistance. In the case of afatinib this profile translates into a robust reduction of colony formation and tumor growth of EGFRG724S-driven cells. Our data provide a mechanistic basis for the osimertinib-induced selection of EGFRG724S-mutant clones and a rationale to treat these patients with clinically approved second-generation EGFR inhibitors.

Comparison of the Detection Rate of Simulated Microcalcifications in Full-Field Digital Mammography, Digital Breast Tomosynthesis, and Synthetically Reconstructed 2-Dimensional Images Performed With 2 Different Digital X-ray Mammography Systems.

OBJECTIVE: The aim of this study was to compare the microcalcification detectability in an anthropomorphic phantom model regarding number, size, and shape in full-field digital mammography (FFDM), synthetically reconstructed 2-dimensional (Synthetic-2D) images, and digital breast tomosynthesis (DBT) performed with 2 different x-ray mammography systems. MATERIALS AND METHODS: Simulated microcalcifications of different numbers (0 to >39), sizes (diameter, 100-800 µm), and shapes (round vs heterogeneous) were scattered by random distribution on 50 film phantoms each divided in 4 quadrants. The FFDM and DBT x-rays were taken from each of these 50 films with both x-ray mammography systems (SenoClaire; GE Healthcare, Selenia Dimensions, Hologic) using an anthropomorphic scattering body and automatic exposure control. The resulting exposure factors were similar to a clinical setting. The synthetically reconstructed 2D images were generated automatically on both systems. All FFDM, Synthetic-2D, and DBT images were interpreted in randomized order and independently of each other by 6 radiologists using a structured questionnaire. RESULTS: The number categories of simulated microcalcifications were correctly evaluated in 55.3% of instances (quadrant by reader) in FFDM, 50.9% in the Synthetic-2D views, and 59.5% in DBT, summarized for 200 quadrants per reader for each Device A and B, respectively. Full-field digital mammography was superior to Synthetic-2D (mean difference, 4%; 95% confidence interval [CI], 2%-7%; P < 0.001), and DBT was superior to both FFDM (mean difference, 4%; 95% CI, 2%-7%; P = 0.002) and Synthetic-2D (mean difference, 9%; 95% CI, 6%-11%; P < 0.001). This trend was consistent in all subgroup analyses. The number of the smallest microcalcifications (100-399 µm) was correctly evaluated in 25.2% of the FFDM, in 14.2% for Synthetic-2D, and in 28.3% of the DBT images. Underestimations of the number of simulated microcalcifications were more common than overestimations. Regarding the size categories of simulated microcalcifications, the rates of correct assessments were in 45.4% of instances in FFDM, 39.9% in the Synthetic-2D views, and 43.6% in DBT, summarized for 200 quadrants per reader and both imaging devices. CONCLUSIONS: In the presented in vitro environment using an anthropomorphic phantom model, standard full-field digital x-ray mammography was superior to synthetically reconstructed 2-dimensional images in the detection of simulated microcalcifications. In view of these results, it is questionable whether Synthetic-2D images can replace FFDM in clinical examinations at the present time. Further investigations are needed to assess the clinical impact of the in vitro results.

Vacuum-assisted breast biopsies (VAB) carried out on an open 1.0T MR imager: Influence of patient and target characteristics on the procedural and clinical results.

PURPOSE: The study was conducted in order to assess the clinical impact of MRI-guided vacuum-assisted breast biopsies carried out using an open 1.0T open MRI-system. MATERIAL AND METHODS: The clinical, imaging, interventional and histological data of all 132 patients with a first MRI-guided vacuum-assisted breast biopsy carried out between 07/2005 and 03/2012at the Radiological Department were extracted from the clinical files. The clinical outcome of patients with benign histological findings was assessed based on the clinical files and queries of the local gynecologists in charge. In the 103 interventional image data sets available target localization and target size were evaluated by two board-certified senior radiologists. Clinical data, lesion characteristics and interventional results were evaluated statistically using subgroup analyses. RESULTS: 131 of 132 MRI-guided breast biopsies (99.2%) were carried out successfully. The median interventional duration was 30min (25%-percentile 25min, 75%-percentile 35min, maximum 75min). Minor complications occurred in 12 interventions of the 131 (9.2%). The histological work-up of the biopsy specimen showed benign results in 98 of 131 interventions (74.8%), lesions with uncertain biological potential in 5 biopsies (3.8%) and malignant findings in 28 biopsies (21.4%). There were 2 false negative histological findings. Neither the patient age nor the medical history nor the anticipated risk of developing breast cancer had an impact on the success rates and the complication rates. In the 103 interventions with available image data sets the maximum target lesion diameters were 1-5mm in 16 lesions (15.5%), 6-10mm in 41 lesions (39.8%) and 11-15mm in 29 lesions (28.2%). There was a positive correlation between the maximum diameters and the rate of malignancy of the target lesions (p=0.020) as well as a trend towards longer interventional procedure durations in smaller target lesions (p=0.183). CONCLUSION: MRI-guided vacuum-assisted breast biopsy for suspicious breast lesions is a clinically safe and feasible method even in small target lesions when using an open high-field MRI-system.

Detection of gastrointestinal bleeding by using multislice computed tomography--acute and chronic hemorrhages.

Nowadays, computed tomography (CT) is established for diagnosing gastrointestinal bleeding. In this retrospective study, the use of CT in diagnosing gastrointestinal bleeding was evaluated. Fifty-three patients received a contrast-medium-enhanced helical multislice CT (MSCT) to locate the bleeding site. Seventy-nine percent of the hemorrhage were acute gastrointestinal bleedings. Fifty-five percent of the acute hemorrhages were located via helical MSCT, 45% of the chronic bleeding sites were detected. Notably, bleeding of diverticula, tumors, and angiodysplasias were well demonstrated. In conclusion, contrast-medium-enhanced MSCT may be used effectively as a noninvasive diagnostic tool for detecting gastrointestinal bleedings.

Optimal systolic and diastolic reconstruction windows for coronary CT angiography using dual-source CT.

OBJECTIVE: The purpose of this study was to determine the position of the optimal systolic and diastolic reconstruction intervals for coronary CT angiography using dual-source CT. SUBJECTS AND METHODS: In 90 patients, coronary dual-source CT angiography was performed without beta-blocking agents. Data were reconstructed in 5% steps throughout the R-R interval. Two independent readers selected optimal systolic and diastolic reconstruction windows for each major coronary vessel--the right coronary artery (RCA), left anterior descending artery (LAD), and left circumflex artery (LCX)--using a 3D viewer and volume-rendering displays. The motion score for each vessel was graded from 1 (no motion artifacts) to 5 (severe motion artifacts over entire vessel). RESULTS: The average heart rate of all patients was 68.7 beats per minute (bpm) (range, 43-119 bpm). The median optimal systolic reconstruction windows were at 35%, 30%, and 35% for the RCA, LAD, and LCX, respectively. The median optimal diastolic reconstruction window was at 75% for all vessels. The mean motion scores (+/- SD) in the systolic reconstructions were 1.9 +/- 0.8 (RCA), 1.7 +/- 0.5 (LAD), and 2.0 +/- 0.6 (LCX). The mean motion scores for the diastolic reconstructions were 1.7 +/- 0.9, 1.5 +/- 0.6, and 1.6 +/- 0.7, respectively. In patients with a heart rate of < 70 bpm, motion scores were significantly lower in diastole versus systole (1.3 +/- 0.4 and 1.9 +/- 0.5, respectively; p < 0.01). In most patients with a heart rate of > 80 bpm, motion scores were lower in systolic than in diastolic reconstructions (2.1 +/- 0.6 and 2.6 +/- 0.8, respectively; p < 0.05). CONCLUSION: Using dual-source CT, the overall optimal reconstruction window is at 75% of the R-R interval in patients with low or intermediate heart rates. In patients with heart rates of > 80 bpm, systolic reconstructions often yield superior image quality compared with diastolic reconstructions.