Lucas L. Geyer

Filtered Back Projection, Adaptive Statistical Iterative Reconstruction, and a Model-based Iterative Reconstruction in Abdominal CT: An Experimental Clinical Study

PURPOSEnTo compare objective and subjective image quality parameters of three image reconstruction algorithms of different generations at routine multidetector computed tomographic (CT) examinations of the abdomen.nnnMATERIALS AND METHODSnThis institutional review board-approved study included 22 consecutive patients (mean age, 56.1 years ± 15.8 [standard deviation]; mean weight, 79.1 kg ± 14.8) who underwent routine CT examinations of the abdomen. A low-contrast phantom was used for objective quality control. Raw data sets were reconstructed by using filtered back projection (FPB), adaptive statistical iterative reconstruction (ASIR), and a model-based iterative reconstruction (MBIR). Radiologists used a semiquantitative scale (-3 to +3) to rate subjective image quality and artifacts, comparing both FBP and MBIR images with ASIR images. The Wilcoxon test and the intraclass correlation coefficient were used to evaluate the data. Measurements of objective noise and CT numbers of soft tissue structures were compared with analysis of variance.nnnRESULTSnThe phantom study revealed an improved detectability of low-contrast targets for MBIR compared with ASIR or FBP. Subjective ratings showed higher image quality for MBIR, with better resolution (median value, 2; range, 1 to 3), lower noise (2; range, 1 to 3), and finer contours (2; range, 1 to 2) compared with ASIR (all P < .001). FBP performed inferiorly (0, range, -2 to 0]; -1 [range, -3 to 0]; 0 [range, -1 to 0], respectively; all, P < .001). Mean interobserver correlation was 0.9 for image perception and 0.7 for artifacts. Objective noise for FBP was 14%-68% higher and for MBIR was 18%-47% lower than that for ASIR (P < .001).nnnCONCLUSIONnThe MBIR algorithm considerably improved objective and subjective image quality parameters of routine abdominal multidetector CT images compared with those of ASIR and FBP.

State of the Art: Iterative CT Reconstruction Techniques

Owing to recent advances in computing power, iterative reconstruction (IR) algorithms have become a clinically viable option in computed tomographic (CT) imaging. Substantial evidence is accumulating about the advantages of IR algorithms over established analytical methods, such as filtered back projection. IR improves image quality through cyclic image processing. Although all available solutions share the common mechanism of artifact reduction and/or potential for radiation dose savings, chiefly due to image noise suppression, the magnitude of these effects depends on the specific IR algorithm. In the first section of this contribution, the technical bases of IR are briefly reviewed and the currently available algorithms released by the major CT manufacturers are described. In the second part, the current status of their clinical implementation is surveyed. Regardless of the applied IR algorithm, the available evidence attests to the substantial potential of IR algorithms for overcoming traditional limitations in CT imaging.

Dynamic CT myocardial perfusion imaging: performance of 3D semi-automated evaluation software.

AbstractObjectivesTo evaluate the performance of three-dimensional semi-automated evaluation software for the assessment of myocardial blood flow (MBF) and blood volume (MBV) at dynamic myocardial perfusion computed tomography (CT).MethodsVolume-based software relying on marginal space learning and probabilistic boosting tree-based contour fitting was applied to CT myocardial perfusion imaging data of 37 subjects. In addition, all image data were analysed manually and both approaches were compared with SPECT findings. Study endpoints included time of analysis and conventional measures of diagnostic accuracy.ResultsOf 592 analysable segments, 42 showed perfusion defects on SPECT. Average analysis times for the manual and software-based approaches were 49.1u2009±u200911.2 and 16.5u2009±u20093.7xa0min respectively (Pu2009<u20090.01). There was strong agreement between the two measures of interest (MBF, ICCu2009=u20090.91, and MBV, ICCu2009=u20090.88, both Pu2009<u20090.01) and no significant difference in MBF/MBV with respect to diagnostic accuracy between the two approaches for both MBF and MBV for manual versus software-based approach; respectively; all comparisons Pu2009>u20090.05.ConclusionsThree-dimensional semi-automated evaluation of dynamic myocardial perfusion CT data provides similar measures and diagnostic accuracy to manual evaluation, albeit with substantially reduced analysis times. This capability may aid the integration of this test into clinical workflows.Key Points• Myocardial perfusion CT is attractive for comprehensive coronary heart disease assessment.n • Traditional image analysis methods are cumbersome and time-consuming.n • Automated 3D perfusion software shortens analysis times.n • Automated 3D perfusion software increases standardisation of myocardial perfusion CT.n • Automated, standardised analysis fosters myocardial perfusion CT integration into clinical practice.

Reduced radiation dose and improved image quality at cardiovascular CT angiography by automated attenuation-based tube voltage selection: intra-individual comparison

AbstractObjectivesTo evaluate the effect of automated tube voltage selection on radiation dose and image quality at cardiovascular CT angiography (CTA).MethodsWe retrospectively analysed paired studies in 72 patients (41 male, 60.5u2009±u200916.5xa0years), who had undergone CTA acquisitions of the heart or aorta both before and after the implementation of an automated x-ray tube voltage selection algorithm (ATVS). All other parameters were kept identical between the two acquisitions. Subjective image quality (IQ) was rated and objective IQ was measured by image noise, signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR) and figure of merit (FOM). Image quality parameters and effective dose were compared between acquisitions.ResultsOverall subjective image quality improved with the percentage of cases scored as adequate or higher increasing from 79xa0% to 92xa0% after implementation of ATVS (Pu2009=u20090.03). SNR (14.1u2009±u20095.9, 15.7u2009±u20096.1, Pu2009=u20090.009), CNR (11.6u2009±u20095.3, 13.2u2009±u20095.6, Pu2009=u20090.011), and FOM (19.9u2009±u200923.3, 43.8u2009±u200951.1, Pu2009<u20090.001) were significantly higher after implementation of ATVS. Mean image noise (24.1u2009±u20098.4 HU, 22.7u2009±u20097.1 HU, Pu2009=u20090.048) and mean effective dose (10.6u2009±u20095.9xa0mSv, 8.8u2009±u20095.0xa0mSv, Pu2009=u20090.003) were significantly lower after implementation of ATVS.ConclusionsAutomated tube voltage selection can operator-independently optimize cardiovascular CTA image acquisition parameters with improved image quality at reduced dose.Key Points• Automatic tube voltage selection optimizes tube voltage for each individual patient.n • In this population, overall radiation dose decreased while image quality improved.n • This tool may become valuable for improving dose/quality ratio.

Iterative image reconstruction: a realistic dose-saving method in cardiac CT imaging?

Iterative techniques are a valuable computed tomography image reconstruction alternative to filtered back projection. In repetitive cycles, iterative algorithms reduce image noise virtually independently of spatial resolution. In light of substantially decreased image noise, tube voltage or current reductions are enabled, resulting in significant radiation dose savings while preserving image quality. Moreover, iterative reconstruction techniques have the advantage of minimizing calcium blooming and metal artifacts. Iterative reconstruction may therefore lead to more exact coronary artery evaluation at constant x-ray tube settings and appears beneficial in clinically challenging scenarios such as overly obese patients, calcified coronary arteries and presence of iatrogenic hardware. For cardiac computed tomography, iterative reconstruction represents a promising and readily available tool.

Time to Recovery of Adrenal Function After Curative Surgery for Cushing's Syndrome Depends on Etiology

CONTEXTnSuccessful tumor resection in endogenous Cushings syndrome (CS) results in tertiary adrenal insufficiency requiring hydrocortisone replacement therapy.nnnOBJECTIVEnThe aim was to analyze the postsurgical duration of adrenal insufficiency of patients with Cushings disease (CD), adrenal CS, and ectopic CS.nnnDESIGNnWe performed a retrospective analysis based on the case records of 230 patients with CS in our tertiary referral center treated from 1983-2014. The mean follow-up time was 8 years.nnnPATIENTSnWe included 91 patients of the three subtypes of CS undergoing curative intended surgery and documented followup after excluding cases with persistent disease, pituitary radiation, concurrent adrenostatic or somatostatin analog treatment, and malignant adrenal disease.nnnRESULTSnThe probability of recovering adrenal function within a 5 years followup differed significantly between subtypes (P = .001). It was 82% in ectopic CS, 58% in CD and 38% in adrenal CS. In the total cohort with restored adrenal function (n = 52) the median time to recovery differed between subtypes: 0.6 years (interquartile range [IQR], 0.03-1.1 y) in ectopic CS, 1.4 years (IQR, 0.9-3.4 y) in CD, and 2.5 years (IQR, 1.6-5.4 y) in adrenal CS (P = .002). In CD the Cox proportional-hazards model showed that the probability of recovery was associated with younger age (hazard ratio, 0.896; 95% confidence interval, 0.822-0.976; P = .012), independently of sex, body mass index, duration of symptoms, and basal ACTH and cortisol levels. There was no correlation with length and extend of hypercortisolism or postoperative glucocorticoid replacement doses.nnnCONCLUSIONSnTime to recovery of adrenal function is dependent on the underlying etiology of CS.

64-MDCT in Mass Casualty Incidents: Volume Image Reading Boosts Radiological Workflow

OBJECTIVEnThe purpose of this study was to evaluate the impact of the use of 64-MDCT and volume image reading on the radiologic workflow during a mass casualty incident simulation.nnnMATERIALS AND METHODSnFor this simulation, casualties were taken to our level I trauma center, and triage was done with whole-body 64-MDCT. The complete raw dataset of thin-section images was sent to a dedicated 3D workstation for further interpretation and simultaneous reformations. This new reading method is called volume image reading. Several time frames were documented to evaluate the workflow: examination time, time needed for image processing, and mean image transfer rates. The results were compared with those of a previous study using a 4-MDCT with axial images only and transfer of data to a PACS.nnnRESULTSnThe time for complete image processing (acquisition, reconstruction, and transfer) for 64-MDCT was 4.1 minutes (range, 3.9-4.3 minutes) compared with 9.0 minutes (range, 6.4-10.2 minutes) for 4-MDCT (p ≤ 0.001). The image processing capacity was 14.8 examinations/h for 64-MDCT compared with 6.7 examinations/h for 4-MDCT. The mean number of images was 953 for 64-MDCT compared with 202 for 4-MDCT (p ≤ 0.001). There were no significant differences between 64- and 4-MDCT for the time needed to prepare patients.nnnCONCLUSIONnThe use of 64-MDCT with volume image reading led to evident advantages in the radiologic trauma workflow compared with 4-MDCT. Reading of the full image set including reformations can be initiated earlier with volume image reading.

CT Evaluation of Small-Diameter Coronary Artery Stents: Effect of an Integrated Circuit Detector with Iterative Reconstruction

PURPOSEnTo use suitable objective methods of analysis to assess the influence of the combination of an integrated-circuit computed tomographic (CT) detector and iterative reconstruction (IR) algorithms on the visualization of small (≤3-mm) coronary artery stents.nnnMATERIALS AND METHODSnBy using a moving heart phantom, 18 data sets obtained from three coronary artery stents with small diameters were investigated. A second-generation dual-source CT system equipped with an integrated-circuit detector was used. Images were reconstructed with filtered back-projection (FBP) and IR at a section thickness of 0.75 mm (FBP75 and IR75, respectively) and IR at a section thickness of 0.50 mm (IR50). Multirow intensity profiles in Hounsfield units were modeled by using a sum-of-Gaussians fit to analyze in-plane image characteristics. Out-of-plane image characteristics were analyzed with z upslope of multicolumn intensity profiles in Hounsfield units. Statistical analysis was conducted with one-way analysis of variance and the Student t test.nnnRESULTSnIndependent of stent diameter and heart rate, IR75 resulted in significantly increased xy sharpness, signal-to-noise ratio, and contrast-to-noise ratio, as well as decreased blurring and noise compared with FBP75 (eg, 2.25-mm stent, 0 beats per minute; xy sharpness, 278.2 vs 252.3; signal-to-noise ratio, 46.6 vs 33.5; contrast-to-noise ratio, 26.0 vs 16.8; blurring, 1.4 vs 1.5; noise, 15.4 vs 21.2; all P < .001). In the z direction, the upslopes were substantially higher in the IR50 reconstructions (2.25-mm stent: IR50, 94.0; IR75, 53.1; and FBP75, 48.1; P < .001).nnnCONCLUSIONnThe implementation of an integrated-circuit CT detector provides substantially sharper out-of-plane resolution of coronary artery stents at 0.5-mm section thickness, while the use of iterative image reconstruction mostly improves in-plane stent visualization.

Low-volume contrast medium protocol for comprehensive cardiac and aortoiliac CT assessment in the context of transcatheter aortic valve replacement.

RATIONALE AND OBJECTIVESnTo investigate the diagnostic performance of a comprehensive computed tomography (CT) protocol for both cardiac and aortoiliac evaluation of patients considered for transcatheter aortic valve replacement (TAVR) using a single, low-volume contrast medium (CM) injection.nnnMATERIALS AND METHODSnForty-four TAVR candidates were retrospectively analyzed. All underwent retrospectively electrocardiogram-gated cardiac CT followed by high-pitch CT angiography of the aortoiliac vasculature using one of two single injection protocols of 320 mgI/mL iodine CM: group A (n = 22), iodine delivery rate-based (1.28 gI/s), 60-mL CM volume, 4.0 mL/s flow rate; group B (n = 22), clinical routine protocol, 100-mL CM volume, 4.0 mL/s flow rate. Mean arterial attenuation, signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR) were calculated. Subjective image quality was assessed.nnnRESULTSnAortic root and iliofemoral dimensions could be analyzed in all cases. Patient characteristics showed no significant differences. Mean attenuation at the levels of the aortic root (285.8 ± 83.0 HU vs 327.5 ± 70.8 HU, P = .080) and the iliofemoral access route (256.8 ± 88.5 HU vs 307.5 ± 93.2 HU, P = .071), as well as SNR and CNR were nonsignificantly lower in group A compared to group B. Subjective image quality was equivalent.nnnCONCLUSIONSnIn multimorbid TAVR patients, the performance of a combined CT protocol using a single low-volume CM bolus is feasible with maintained image quality compared to a standard protocol.

Automated Quantification of Epicardial Adipose Tissue Using CT Angiography: Evaluation of a Prototype Software

ObjectivesThis study evaluated the performance of a novel automated software tool for epicardial fat volume (EFV) quantification compared to a standard manual technique at coronary CT angiography (cCTA).MethodscCTA data sets of 70 patients (58.6u2009±u200912.9xa0years, 33 men) were retrospectively analysed using two different post-processing software applications. Observer 1 performed a manual single-plane pericardial border definition and EFVM segmentation (manual approach). Two observers used a software program with fully automated 3D pericardial border definition and EFVA calculation (automated approach). EFV and time required for measuring EFV (including software processing time and manual optimization time) for each method were recorded. Intraobserver and interobserver reliability was assessed on the prototype software measurements. T test, Spearman’s rho, and Bland–Altman plots were used for statistical analysis.ResultsThe final EFVA (with manual border optimization) was strongly correlated with the manual axial segmentation measurement (60.9u2009±u200933.2xa0mL vs. 65.8u2009±u200937.0xa0mL, rhou2009=u20090.970, Pu2009<u20090.001). A mean of 3.9u2009±u20091.9 manual border edits were performed to optimize the automated process. The software prototype required significantly less time to perform the measurements (135.6u2009±u200924.6xa0s vs. 314.3u2009±u200976.3xa0s, Pu2009<u20090.001) and showed high reliability (ICCu2009>u20090.9).ConclusionsAutomated EFVA quantification is an accurate and time-saving method for quantification of EFV compared to established manual axial segmentation methods.Key Points• Manual epicardial fat volume quantification correlates with risk factors but is time-consuming.• The novel software prototype automates measurement of epicardial fat volume with good accuracy.• This novel approach is less time-consuming and could be incorporated into clinical workflow.