Model-based iterative reconstruction for reduction of radiation dose in abdominopelvic CT: comparison to adaptive statistical iterative reconstruction

Reduction of metal artefacts from bilateral hip prostheses during lower extremity computed tomography angiography: an experimental phantom study

INTRODUCTION

Image quality reduction due to metallic artefacts is a significant challenge during vascular computed tomography (CT) imaging of the lower extremities in patients with hip prostheses. This study aims to analyse various reconstruction algorithms' ability to reduce metal artefacts due to two types of hip prostheses during lower extremity CT angiography examinations.

METHODS

A pelvis phantom was fabricated with the insertion of a tube filled with contrast media to simulate the femoral artery, and the phantom was then CT scanned with and without hip prostheses. Multimodal images were acquired using different kilovoltage peak (kVp) settings and reconstructed with different algorithms, such as filtered back projection (FBP), iterative reconstruction (iDose4), iterative model-based reconstruction (IMR) and orthopaedic metal artefact reduction (O-MAR). Image quality was assessed based on image noise, signal-to-noise ratio (SNR) and Hounsfield unit (HU) deviation.

RESULTS

The IMR approach significantly improved image quality compared to iDose4 and FBP. For the vascular region, O-MAR improves SNR by 5 ± 1, 23 ± 5 and 42 ± 9 for FBP, iDose4 and IMR respectively, and improves HU precision towards the baseline values by 49% and 83% for FBP and IMR, respectively. The noise reduction was 71% and 89% for FBP and IMR, and 57% for iDose4. O-MAR greatly enhances SNR corrections among the most severe artefacts, with 29 ± 1 and 43 ± 4 for FBP and IMR, compared to iDose4 by 37 ± 7.

CONCLUSION

IMR combined with O-MAR could improve the CT angiography of the lower extremities of patients with a hip prosthesis.

Detection of Vertebral Mass and Diagnosis of Spinal Cord Compression in Computed Tomography With Deep Learning Reconstruction: Comparison With Hybrid Iterative Reconstruction

PURPOSE

To compare the impact of deep learning reconstruction (DLR) and hybrid-iterative reconstruction (hybrid-IR) on vertebral mass depiction, detection, and diagnosis of spinal cord compression on computed tomography (CT).

METHODS

This retrospective study included 29 and 20 patients with and without vertebral masses. CT images were reconstructed using DLR and hybrid-IR. Three readers performed vertebral mass detection tests and evaluated the presence of spinal cord compression, the depiction of vertebral masses, and image noise. Quantitative image noise was measured by placing regions of interest on the aorta and spinal cord.

RESULTS

Deep learning reconstruction tended to improve the performance of readers with less diagnostic imaging experience in detecting vertebral masses (area under the receiver operating characteristic curve [AUC] = .892-.966) compared with hybrid-IR (AUC = .839-.917). Diagnostic performance in evaluating spinal cord compression in DLR (AUC = .887-.995) also improved compared with that in hybrid-IR (AUC = .866-.942) for some readers. The depiction of vertebral masses and subjective image noise in DLR were significantly improved compared with those in hybrid-IR (P < .041). Quantitative image noise in DLR was also significantly reduced compared with that in hybrid-IR (P < .001).

CONCLUSION

Deep learning reconstruction improved the depiction of vertebral masses, which resulted in a tendency to improve the performance of CT compared to hybrid-IR in detecting vertebral masses and diagnosing spinal cord compression for some readers.

Engineering and clinical use of artificial intelligence (AI) with machine learning and data science advancements: radiology leading the way for future

Over the years, many clinical and engineering methods have been adapted for testing and screening for the presence of diseases. The most commonly used methods for diagnosis and analysis are computed tomography (CT) and X-ray imaging. Manual interpretation of these images is the current gold standard but can be subject to human error, is tedious, and is time-consuming. To improve efficiency and productivity, incorporating machine learning (ML) and deep learning (DL) algorithms could expedite the process. This article aims to review the role of artificial intelligence (AI) and its contribution to data science as well as various learning algorithms in radiology. We will analyze and explore the potential applications in image interpretation and radiological advances for AI. Furthermore, we will discuss the usage, methodology implemented, future of these concepts in radiology, and their limitations and challenges.

Scanner dependence of adaptive statistical iterative reconstruction with 3D noise power spectrum central frequency and noise magnitude ratios

PURPOSE

In this study, the noise reduction properties of the adaptive statistical iterative reconstruction (IR) on two different CT scanners of 64 and 256-slice were compared and their differences were assessed.

METHODS AND MATERIALS

The homogeneous module of the ACR CT phantom was scanned on the 64 and 256 slices CT scanners from the same vendor in the range of 15-40 mA. On each scanner, the data were reconstructed using filtered back projection (FBP) and at all strengths of IR with the STANDARD kernel. For each reconstruction, a 3D noise power spectrum (NPS) was calculated and the central frequency ratio in the xy plane (CFR_xy ), CFR in the z-direction (CFR_z ), and noise magnitude ratio (NMR) were derived. CFR is the central frequency ratio of NPS between the denoised image and the FBP image, and NMR is the ratio of the areas under the NPS curves. Ideally, both CFR_xy and CFR_z should be near 1, indicating minimal texture changes in both xy and z directions, while NMR should be as close to 0 as possible, indicating more noise reduction.

RESULTS

When comparing strengths with equivalent impact on noise texture, IR on the 64-slice reduced the noise magnitude in the xy plane more than that on the 256-slice. In the z-direction, the IR on the 256-slice produced a central frequency shift on the 256-slice but not on the 64-slice. In addition, the noise reduction effects of the IR on the 256-slice were affected when radiation exposure was below 2.0 mGy, but there was no observable dose-dependence on the 64-slice.

CONCLUSIONS

Our noise property analysis revealed that iterative reconstructions on different scanner platforms from the same vendor can be distinct, with unique effects on the noise texture and magnitude in CT images. The IR on a 64-slice scanner provides slightly enhanced noise reduction and maintains a noise reduction rate independent of dose, unlike the one on a 256-slice scanner. Notably, the IR on the 64-slice scanner was a 2D noise reduction technique (NRT), while the one on the 256-slice was a 3D NRT. These observations showcase the impact of different NRTs on clinical CT images, even when comparing the same NRT on different scanners.

Factors Affecting Radiation Dose in Computed Tomography Angiograms for Pulmonary Embolism: A Retrospective Cohort Study

Objectives

Computed tomography pulmonary angiogram (CTPA) is one of the most commonly ordered and frequently overused tests. The purpose of this study was to evaluate the mean radiation dose to patients getting CTPA and to identify factors that are associated with higher dose.

Material and Methods

This institutionally approved retrospective study included all patients who had a CTPA to rule out acute pulmonary embolism between 2016 and 2018 in a tertiary care center. Patient data (age, sex, body mass index [BMI], and patient location), CT scanner type, image reconstruction methodology, and radiation dose parameters (dose-length product [DLP]) were recorded. Effective dose estimates were obtained by multiplying DLP by conversion coefficient (0.014 mSv•mGy^-1•cm^-1). Multivariate logistic regression analysis was performed to determine the factors affecting the radiation dose.

Results

There were 2342 patients (1099 men and 1243 women) with a mean age of 58.1 years (range 0.2-104.4 years) and BMI of 31.3 kg/m² (range 12-91.5 kg/m²). The mean effective radiation dose was 5.512 mSv (median - 4.27 mSv; range 0.1-43.0 mSv). Patient factors, including BMI >25 kg/m², male sex, age >18 years, and intensive care unit (ICU) location, were associated with significantly higher dose (P < 0.05). CT scanning using third generation dual-source scanner with model-based iterative reconstruction (IR) had significantly lower dose (mean: 4.90 mSv) versus single-source (64-slice) scanner with filtered back projection (mean: 9.29 mSv, P < 0.001).

Conclusion

Patients with high BMI and ICU referrals are associated with high CT radiation dose. They are most likely to benefit by scanning on newer generation scanner using advance model-based IR techniques.

Effect of a New Model-Based Reconstruction Algorithm for Evaluating Early Peripheral Lung Cancer With Submillisievert Chest Computed Tomography

OBJECTIVE

The aim of this study was to compare a new model-based iterative reconstruction algorithm with either spatial and density resolution balance (MBIRSTND) or spatial resolution preference (MBIRRP20) with the adaptive statistical iterative reconstruction (ASIR) in evaluating early small peripheral lung cancer (SPLC) with submillisievert chest computed tomography (CT).

METHODS

Low-contrast and spatial resolutions were assessed in a phantom and with 30 pathologically confirmed SPLC patients. Images were reconstructed using 40% ASIR, MBIRSTND, and MBIRRP20. Computed tomography value and image noise were measured by placing the regions of interest on back muscle and subcutaneous fat at 3 levels. Two radiologists used a 4-point scale (1, worst, and 4, best) to rate subjective image quality in 3 aspects: image noise, nodule imaging signs, and nodule internal clarity.

RESULTS

The phantom study revealed an improved detectability of low-contrast targets and small objects for MBIRSTND and MBIRRP20 compared with ASIR. The effective dose for patient scans was 0.88 ± 0.83 mSv. There was no significant difference in CT value between the 3 reconstructions (P > 0.05), but MBIRSTND and MBIRRP20 significantly reduced image noise compared with ASIR (P < 0.05): 15.69 ± 1.83 HU and 29.97 ± 3.84 HU versus 51.06 ± 11.02 HU in the back muscle, and 15.96 ± 3.07 HU and 27.37 ± 3.88 HU versus 38.04 ± 8.87 HU in subcutaneous fat, respectively. Among the 3 reconstructions, MBIRSTND was the best in reducing image noise and identifying the internal compositions of cancer nodules, and MBIRRP20 was the best in analyzing the internal and external signs of pulmonary nodules.

CONCLUSIONS

Submillisievert chest CT reconstructed with MBIRSTND and MBIRRP20 provides superior images for the detailed analyses of SPLC compared with ASIR.

Influence of slice thickness on result of CT histogram analysis in indeterminate adrenal masses

PURPOSE

The aim was to determine the optimal slice thickness of CT images and the optimal threshold of negative voxels for CT histogram analysis to distinguish adrenal adenomas from non-adenomas with a mean attenuation more than 10 Hounsfield units (HU).

METHODS

Volume CT histogram analysis of 83 lipid-poor adenomas and 80 non-adenomas was performed retrospectively. The volume of interest was extracted from each adrenal lesion, and the mean attenuation, standard deviation (SD), and percentage of voxels with a negative CT value were recorded using reconstructions with different slice thicknesses (5 mm, 2.5 mm, 1.25 mm). The percentage of negative voxels was correlated with SD as a measure of image noise and with the reference splenic tissue values. The sensitivity, specificity, and positive predictive value (PPV) for the identification of adenomas were calculated using reconstructions with different slice thicknesses and three different thresholds of negative voxels (1%, 5%, 10%).

RESULTS

The percentage of negative voxels increased with a thinner slice thickness and correlated with increasing CT image noise in adenomas, non-adenomas, and spleen. Using a threshold of 10% negative voxels and a slice thickness of 5 mm, we reached a sensitivity of 53.0%, specificity of 98.8% and the highest PPV, and thus we propose this combination for clinical use. Other combinations achieved a clearly lower specificity and PPV as a result of the increasing noise in CT images.

CONCLUSION

The CT slice thickness significantly affects the result and diagnostic value of histogram analysis. Thin CT slice reconstructions are inappropriate for histogram analysis.

CT radiation dose reduction in patients with total hip arthroplasties using model-based iterative reconstruction and orthopaedic metal artefact reduction

OBJECTIVE

To evaluate the impact of radiation dose reduction on image quality in patients with metal-on-metal total hip arthroplasties (THAs) using model-based iterative reconstruction (MBIR) combined with orthopaedic metal artefact reduction (O-MAR).

MATERIALS AND METHODS

Patients with metal-on-metal THAs received a pelvic CT with a full (FD) and a reduced radiation dose (RD) with -20%, -40%, -57%, or -80% CT radiation dose respectively, when assigned to group 1, 2, 3, or 4 respectively. FD acquisitions were reconstructed with iterative reconstruction, iDose⁴. RD acquisitions were additionally reconstructed with iterative model-based reconstruction (IMR) levels 1-3 with different levels of noise suppression. CT numbers, noise and contrast-to-noise ratios were measured in muscle, fat and bladder. Subjective image quality was evaluated on seven aspects including artefacts, osseous structures, prosthetic components and soft tissues.

RESULTS

Seventy-six patients were randomly assigned to one of the four groups. While reducing radiation dose by 20%, 40%, 57%, or 80% in combination with IMR, CT numbers remained constant. Compared with iDose⁴, the noise decreased (p < 0.001) and contrast-to-noise ratios increased (p < 0.001) with IMR. O-MAR improved CT number accuracy in the bladder and reduced noise in the bladder, muscle and fat (p < 0.01). Subjective image quality was rated lower on RD IMR images than FD iDose⁴ images on all seven aspects (p < 0.05) and was not related to the applied radiation dose reduction.

CONCLUSION

In RD IMR with O-MAR images, CT numbers remained constant, noise decreased and contrast-to-noise ratios between muscle and fat increased compared with FD iDose⁴ with O-MAR images in patients with metal-on-metal THAs. Subjective image quality reduced, regardless of the degree of radiation dose reduction.

Comparison of image noise and image quality between full-dose abdominal computed tomography scans reconstructed with weighted filtered back projection and half-dose scans reconstructed with improved sinogram-affirmed iterative reconstruction (SAFIRE*)

PURPOSE

To retrospectively compare the image noise, signal-to-noise ratio (SNR), and subjective image quality between CT images acquired with a dual-source, split-dose imaging protocol reconstructed at full and half doses with weighted filtered back projection (wFBP) and an improved sinogram-affirmed iterative reconstruction algorithm (SAFIRE*).

METHODS

Fifty-three consecutive patients underwent contrast-enhanced CT of the abdomen using a standardized dual-source, single energy CT protocol. Half-dose images were retrospectively generated using data from one detector only. Full-dose datasets were reconstructed with wFBP, while half-dose datasets were reconstructed with wFBP and SAFIRE* strengths 1-5. Region of interest analysis was performed to assess SNR and noise. Diagnostic acceptability, subjective noise, and spatial resolution were graded on a 10-point scale by two readers. Statistical analysis was carried out with repeated measures analysis of variance, Wilcoxon signed rank test, and Cohen's κ test.

RESULTS

With the increasing strengths of SAFIRE*, a progressive reduction in noise and increase in SNR (p < 0.01) was observed. There was a statistically significant decrease in objective noise and increase in SNR in half-dose SAFIRE* strength 4 and 5 reconstructions compared to full-dose reconstructions using wFBP (p < 0.01). Qualitative analysis revealed a progressive increase in diagnostic acceptability, decrease in subjective noise and increase in spatial resolution for half-dose images reconstructed with the increasing strengths of SAFIRE* (p < 0.01).

CONCLUSIONS

Half-dose CT images reconstructed with SAFIRE* at strength 4 and 5 have superior image quality compared to full-dose images reconstructed with wFBP. SAFIRE* potentially allows dose reductions in the order of 50% over wFBP.

Contrast-Enhanced CT with Knowledge-Based Iterative Model Reconstruction for the Evaluation of Parotid Gland Tumors: A Feasibility Study

Objective

The purpose of this study was to determine the diagnostic utility of low-dose CT with knowledge-based iterative model reconstruction (IMR) for the evaluation of parotid gland tumors.

Materials and Methods

This prospective study included 42 consecutive patients who had undergone low-dose contrast-enhanced CT for the evaluation of suspected parotid gland tumors. Prior or subsequent non-low-dose CT scans within 12 months were available in 10 of the participants. Background noise (BN), signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR) were compared between non-low-dose CT images and images generated using filtered back projection (FBP), hybrid iterative reconstruction (iDose⁴; Philips Healthcare), and knowledge-based IMR. Subjective image quality was rated by two radiologists using five-point grading scales to assess the overall image quality, delineation of lesion contour, image sharpness, and noise.

Results

With the IMR algorithm, background noise (IMR, 4.24 ± 3.77; iDose⁴, 8.77 ± 3.85; FBP, 11.73 ± 4.06; p = 0.037 [IMR vs. iDose⁴] and p < 0.001 [IMR vs. FBP]) was significantly lower and SNR (IMR, 23.93 ± 7.49; iDose⁴, 10.20 ± 3.29; FBP, 7.33 ± 2.03; p = 0.011 [IMR vs. iDose⁴] and p < 0.001 [IMR vs. FBP]) was significantly higher compared with the other two algorithms. The CNR was also significantly higher with the IMR compared with the FBP (25.76 ± 11.88 vs. 9.02 ± 3.18, p < 0.001). There was no significant difference in BN, SNR, and CNR between low-dose CT with the IMR algorithm and non-low-dose CT. Subjective image analysis revealed that IMR-generated low-dose CT images showed significantly better overall image quality and delineation of lesion contour with lesser noise, compared with those generated using FBP by both reviewers 1 and 2 (4 vs. 3; 4 vs. 3; and 3–4 vs. 2; p < 0.05 for all pairs), although there was no significant difference in subjective image quality scores between IMR-generated low-dose CT and non-low-dose CT images.

Conclusion

Iterative model reconstruction-generated low-dose CT is an alternative to standard non-low-dose CT without significantly affecting image quality for the evaluation of parotid gland tumors.

Forward-Projected Model-Based Iterative Reconstruction in Screening Low-Dose Chest CT: Comparison With Adaptive Iterative Dose Reduction 3D

OBJECTIVE

The objective of this study is to compare forward-projected model-based iterative reconstruction solution (FIRST), a newer fully iterative CT reconstruction method, with adaptive iterative dose reduction 3D (AIDR 3D) in low-dose screening CT for lung cancer. Differences in image noise, image quality, and pulmonary nodule detection, size, and characterization were specifically evaluated.

MATERIALS AND METHODS

Low-dose chest CT images obtained for 50 consecutive patients between December 2015 and January 2016 were retrospectively reviewed. Images were reconstructed using FIRST and AIDR 3D for both lung and soft-tissue reconstruction. Images were independently reviewed to assess image noise, subjective image quality (with use of a 5-point Likert scale, with 1 denoting far superior image quality; 2, superior quality; 3, equivalent quality; 4, inferior quality; and 5, far inferior quality), pulmonary nodule count, size of the largest pulmonary nodule, and characterization of the largest pulmonary nodule (i.e., solid, part solid, or ground glass).

RESULTS

Across all 50 cases, measured image noise was lower with FIRST than with AIDR 3D (lung window, 44% reduction, 41 ± 7 vs 74 ± 8 HU, respectively; soft-tissue window, 32% reduction, 11 ± 2 vs 16 ± 2 HU, respectively). Readers subjectively rated images obtained with FIRST as comparable to images obtained with AIDR 3D (mean [± SD] Likert score for FIRST vs AIDR 3D, 3.2 ± 0.3 for soft-tissue reconstructions and 3.0 ± 0.3 for lung reconstructions). For each reader, very good agreement regarding nodule count was noted between FIRST and AIDR 3D (interclass correlation coefficient [ICC], 0.83 for reader 1 and 0.78 for reader 2). Excellent agreement regarding nodule size (ICC, 0.99 for reader 1 and 0.99 for reader 2) and characterization of the largest nodule (kappa value, 0.92 for reader 1 and 0.82 for reader 2) also existed.

CONCLUSION

Images reconstructed with FIRST are superior to those reconstructed AIDR 3D with regard to image noise and are equivalent with regard to subjective image quality, pulmonary nodule count, and nodule characterization.

Adaptive statistical iterative reconstruction use for radiation dose reduction in pediatric lower-extremity CT: impact on diagnostic image quality

BACKGROUND

For the past several years, increased levels of imaging radiation and cumulative radiation to children has been a significant concern. Although several measures have been taken to reduce radiation dose during computed tomography (CT) scan, the newer dose reduction software adaptive statistical iterative reconstruction (ASIR) has been an effective technique in reducing radiation dose. To our knowledge, no studies are published that assess the effect of ASIR on extremity CT scans in children.

OBJECTIVE

To compare radiation dose, image noise, and subjective image quality in pediatric lower extremity CT scans acquired with and without ASIR.

MATERIALS AND METHODS

The study group consisted of 53 patients imaged on a CT scanner equipped with ASIR software. The control group consisted of 37 patients whose CT images were acquired without ASIR. Image noise, Computed Tomography Dose Index (CTDI) and dose length product (DLP) were measured. Two pediatric radiologists rated the studies in subjective categories: image sharpness, noise, diagnostic acceptability, and artifacts.

RESULTS

The CTDI (p value = 0.0184) and DLP (p value <0.0002) were significantly decreased with the use of ASIR compared with non-ASIR studies. However, the subjective ratings for sharpness (p < 0.0001) and diagnostic acceptability of the ASIR images (p < 0.0128) were decreased compared with standard, non-ASIR CT studies.

CONCLUSION

Adaptive statistical iterative reconstruction reduces radiation dose for lower extremity CTs in children, but at the expense of diagnostic imaging quality. Further studies are warranted to determine the specific utility of ASIR for pediatric musculoskeletal CT imaging.

Multidetector Computed Tomography Imaging: Effect of Sparse Sampling and Iterative Reconstruction on Trabecular Bone Microstructure

Multidetector computed tomography-based trabecular bone microstructure analysis ensures promising results in fracture risk prediction caused by osteoporosis. Because multidetector computed tomography is associated with high radiation exposure, its clinical routine use is limited. Hence, in this study, we investigated in 11 thoracic midvertebral specimens whether trabecular texture parameters are comparable derived from (1) images reconstructed using statistical iterative reconstruction (SIR) and filtered back projection as criterion standard at different exposures (80, 150, 220, and 500 mAs) and (2) from SIR-based sparse sampling projections (12.5%, 25%, 50%, and 100%) and equivalent exposures as criterion standard. Twenty-four texture features were computed, and those that showed similar values between (1) filtered back projection and SIR at the different exposure levels and (2) sparse sampling and equivalent exposures and reconstructed with SIR were identified. These parameters can be of equal value in determining trabecular bone microstructure with lower radiation exposure using sparse sampling and SIR.

Image quality and radiation dose of CT venography with double dose reduction using model based iterative reconstruction: comparison with conventional CT venography using filtered back projection

Background Computed tomography venography (CTV) at low kVp using model-based iterative reconstruction (MBIR) can enhance vascular enhancement with noise reduction. Purpose To evaluate image qualities and radiation doses of CTV at 80 kVp using MBIR and a small iodine contrast media (CM) dose and to compare these with those of CTV performed using a conventional protocol. Material and Methods Sixty-five patients (mean age = 58.1 ± 7.2 years) that underwent CTV for the evaluation of deep vein thrombosis (DVT) and varicose veins were enrolled in this study. Patients were divided into two groups: Group A (35 patients, 80 kVp, MBIR, automatic tube current modulation, CM = 270 mg/mL, 100 mL) and Group B (30 patients, 100 kVp, filtered back projection [FBP], 120 fixed mA, CM = 370 mg/mL, 120 mL). Objective and subjective image qualities of inferior vena cava (IVC), femoral vein (FV), and popliteal vein (PV) were assessed and radiation doses were recorded. Results Mean vascular enhancement in group A was significantly lower than in group B ( P < 0.01). Noise in group A was significantly lower than in group B except for PV and contrast-to-noise ratio were not significantly different in the two groups ( P > 0.05). In addition, radiation dose in group A was significantly lower than in group B ( P < 0.001). Subjective image quality comparison revealed group A was statistically inferior to group B except for subjective image noise. Conclusion CTV at 80 kVp using MBIR with small iodine contrast dose provided acceptable image quality at a lower radiation dose than conventional CTV using FBP.

Deep learning with convolutional neural network in radiology

Deep learning with a convolutional neural network (CNN) is gaining attention recently for its high performance in image recognition. Images themselves can be utilized in a learning process with this technique, and feature extraction in advance of the learning process is not required. Important features can be automatically learned. Thanks to the development of hardware and software in addition to techniques regarding deep learning, application of this technique to radiological images for predicting clinically useful information, such as the detection and the evaluation of lesions, etc., are beginning to be investigated. This article illustrates basic technical knowledge regarding deep learning with CNNs along the actual course (collecting data, implementing CNNs, and training and testing phases). Pitfalls regarding this technique and how to manage them are also illustrated. We also described some advanced topics of deep learning, results of recent clinical studies, and the future directions of clinical application of deep learning techniques.

Iterative Reconstruction Designed for Brain CT: A Correlative Study With Filtered Back Projection for the Diagnosis of Acute Ischemic Stroke

OBJECTIVES

The objective of this study is to evaluate the usefulness of iterative model reconstruction designed for brain computed tomography (CT) (IMR-Neuro) for the diagnosis of acute ischemic stroke.

METHODS

This retrospective study included 20 patients with acute middle cerebral artery infarction who have undergone brain CT and 20 nonstroke patients (control). We reconstructed axial images with filtered back projection (FBP) and IMR-Neuro (slice thickness, 1 and 5 mm). We compared the CT number of the infarcted area, the image noise, contrast, and the contrast to noise ratio of the infarcted and the noninfarcted areas between the different reconstruction methods. We compared the performance of 10 radiologists in the detection of parenchymal hypoattenuation between 2 techniques using the receiver operating characteristic (ROC) techniques with the jackknife method.

RESULTS

The image noise was significantly lower with IMR-Neuro [5 mm: 2.5 Hounsfield units (HU) ± 0.5, 1 mm: 3.9 HU ± 0.5] than with FBP (5 mm: 4.9 HU ± 0.5, 1 mm: 10.1 HU ± 1.4) (P < 0.01). The contrast to noise ratio was significantly greater with IMR-Neuro (5 mm: 2.6 ± 2.1, 1 mm: 1.6 ± 1.3) than with FBP (5 mm: 1.2 ± 1.0; 1 mm: 0.6 ± 0.5) (P < 0.01). The value of the average area under the receiver operating curve was significantly higher with IMR-Neuro than FBP (5 mm: 0.79 vs 0.74, P = 0.04; 1 mm: 0.76 vs 0.69, P = 0.04).

CONCLUSIONS

Compared with FBP, IMR-Neuro improves the image quality and the performance for the detection of parenchymal hypoattenuation with acute ischemic stroke.

Machine Learning in Radiology: Applications Beyond Image Interpretation

Much attention has been given to machine learning and its perceived impact in radiology, particularly in light of recent success with image classification in international competitions. However, machine learning is likely to impact radiology outside of image interpretation long before a fully functional "machine radiologist" is implemented in practice. Here, we describe an overview of machine learning, its application to radiology and other domains, and many cases of use that do not involve image interpretation. We hope that better understanding of these potential applications will help radiology practices prepare for the future and realize performance improvement and efficiency gains.

Reduced dose CT with model-based iterative reconstruction compared to standard dose CT of the chest, abdomen, and pelvis in oncology patients: intra-individual comparison study on image quality and lesion conspicuity

PURPOSE

To compare image quality and lesion conspicuity of reduced dose (RD) CT with model-based iterative reconstruction (MBIR) compared to standard dose (SD) CT in patients undergoing oncological follow-up imaging.

METHODS

Forty-four cancer patients who had a staging SD CT within 12 months were prospectively included to undergo a weight-based RD CT with MBIR. Radiation dose was recorded and tissue attenuation and image noise of four tissue types were measured. Reproducibility of target lesion size measurements of up to 5 target lesions per patient were analyzed. Subjective image quality was evaluated for three readers independently utilizing 4- or 5-point Likert scales.

RESULTS

Median radiation dose reduction was 46% using RD CT (P < 0.01). Median image noise across all measured tissue types was lower (P < 0.01) in RD CT. Subjective image quality for RD CT was higher (P < 0.01) in regard to image noise and overall image quality; however, there was no statistically significant difference regarding image sharpness (P = 0.59). There were subjectively more artifacts on RD CT (P < 0.01). Lesion conspicuity was subjectively better in RD CT (P < 0.01). Repeated target lesion size measurements were highly reproducible both on SD CT (ICC = 0.987) and RD CT (ICC = 0.97).

CONCLUSIONS

RD CT imaging with MBIR provides diagnostic imaging quality and comparable lesion conspicuity on follow-up exams while allowing dose reduction by a median of 46% compared to SD CT imaging.

Full and hybrid iterative reconstruction to reduce artifacts in abdominal CT for patients scanned without arm elevation

Background Abdominal computed tomography (CT) without arm elevation is associated with degraded image quality due to streak artifacts. Purpose To compare the degree of streak artifacts in abdominal CT images without arm elevation between full iterative reconstruction (IR), hybrid IR, and filtered back projection (FBP) using two commercially available scanners. Material and Methods First, a phantom study simulating CT examination without arm elevation was performed. Second, unenhanced axial images of 33 patients (17 and 16 patients for each vendor) who underwent CT without arm elevation were reconstructed with full IR, hybrid IR and FBP. A radiologist placed 50 parallel lines with lengths of 50 pixels vertical to the streaks and quantitatively evaluated the images for streak artifacts in the phantom study. Two radiologists evaluated the images of patients for streak artifacts (on the liver and the kidney) and diagnostic acceptability using a four-point scale. Results The phantom study indicated that full IR algorithms were more effective than FBP in reducing streak artifacts. In the clinical patient study, streak artifacts were significantly more reduced with full IR compared with FBP in both the liver and kidney ( P < 0.012). Streak artifact reduction was limited with hybrid IR. Model-based iterative reconstruction (MBIR) (one of the full IR algorithms) provided diagnostically more acceptable image quality ( P < 0.016) compared with FBP. Conclusion In abdominal CT without arm elevation, full IR enabled a more efficient streak artifact reduction compared with FBP and MBIR was associated with diagnostically more acceptable images.

Implementation of ultra-low-dose lung protocols in CT-guided lung biopsies: feasibility and safety in the clinical setting

Objective

To evaluate the use of ultra-low-dose computed tomography (ULDCT) for CT-guided lung biopsy versus standard-dose CT (SDCT).

Methods

CT-guided lung biopsies from 115 patients (50 ULDCT, 65 SDCT) were analyzed retrospectively. SDCT settings were 120 kVp with automatic mAs modulation. ULDCT settings were 80 kVp with fixed exposure (20 mAs). Two radiologists evaluated image quality (i.e., needle artifacts, lesion contouring, vessel recognition, visibility of interlobar fissures). Complications and histological results were also evaluated.

Results

ULDCT was considered feasible for all lung interventions, showing the same diagnostic accuracy as SDCT. Its mean total radiation dose (dose–length product) was significantly reduced to 34 mGy-cm (SDCT 426 mGy-cm). Image quality and complication rates (P = 0.469) were consistent.

Conclusions

ULDCT for CT-guided lung biopsies appears safe and accurate, with a significantly reduced radiation dose. We therefore recommend routine clinical use of ULDCT for the benefit of patients and interventionalists.

Improvements to image quality using hybrid and model-based iterative reconstructions: a phantom study

BACKGROUND

The number of computed tomography (CT) examinations is increasing and leading to an increase in total patient exposure. It is therefore important to optimize CT scan imaging conditions in order to reduce the radiation dose. The introduction of iterative reconstruction methods has enabled an improvement in image quality and a reduction in radiation dose.

PURPOSE

To investigate how image quality depends on reconstruction method and to discuss patient dose reduction resulting from the use of hybrid and model-based iterative reconstruction.

MATERIAL AND METHODS

An image quality phantom (Catphan® 600) and an anthropomorphic torso phantom were examined on a Philips Brilliance iCT. The image quality was evaluated in terms of CT numbers, noise, noise power spectra (NPS), contrast-to-noise ratio (CNR), low-contrast resolution, and spatial resolution for different scan parameters and dose levels. The images were reconstructed using filtered back projection (FBP) and different settings of hybrid (iDose⁴) and model-based (IMR) iterative reconstruction methods.

RESULTS

iDose⁴ decreased the noise by 15-45% compared with FBP depending on the level of iDose⁴. The IMR reduced the noise even further, by 60-75% compared to FBP. The results are independent of dose. The NPS showed changes in the noise distribution for different reconstruction methods. The low-contrast resolution and CNR were improved with iDose⁴, and the improvement was even greater with IMR.

CONCLUSION

There is great potential to reduce noise and thereby improve image quality by using hybrid or, in particular, model-based iterative reconstruction methods, or to lower radiation dose and maintain image quality.

Submillisievert Radiation Dose Coronary CT Angiography: Clinical Impact of the Knowledge-Based Iterative Model Reconstruction

RATIONALE AND OBJECTIVES

The purpose of this study was to evaluate the noise and image quality of images reconstructed with a knowledge-based iterative model reconstruction (knowledge-based IMR) in ultra-low dose cardiac computed tomography (CT).

MATERIALS AND METHODS

We performed submillisievert radiation dose coronary CT angiography on 43 patients. We also performed a phantom study to evaluate the influence of object size with the automatic exposure control phantom. We reconstructed clinical and phantom studies with filtered back projection (FBP), hybrid iterative reconstruction (hybrid IR), and knowledge-based IMR. We measured effective dose of patients and compared CT number, image noise, and contrast noise ratio in ascending aorta of each reconstruction technique. We compared the relationship between image noise and body mass index for the clinical study, and object size for phantom study.

RESULTS

The mean effective dose was 0.98 ± 0.25 mSv. The image noise of knowledge-based IMR images was significantly lower than those of FBP and hybrid IR images (knowledge-based IMR: 19.4 ± 2.8; FBP: 126.7 ± 35.0; hybrid IR: 48.8 ± 12.8, respectively) (P < .01). The contrast noise ratio of knowledge-based IMR images was significantly higher than those of FBP and hybrid IR images (knowledge-based IMR: 29.1 ± 5.4; FBP: 4.6 ± 1.3; hybrid IR: 13.1 ± 3.5, respectively) (P < .01). There were moderate correlations between image noise and body mass index in FBP (r = 0.57, P < .01) and hybrid IR techniques (r = 0.42, P < .01); however, these correlations were weak in knowledge-based IMR (r = 0.27, P < .01).

CONCLUSION

Compared to FBP and hybrid IR, the knowledge-based IMR offers significant noise reduction and improvement in image quality in submillisievert radiation dose cardiac CT.

Accurate tissue characterization in low-dose CT imaging with pure iterative reconstruction

INTRODUCTION

We assess the ability of low-dose hybrid iterative reconstruction (IR) and 'pure' model-based IR (MBIR) images to maintain accurate Hounsfield unit (HU)-determined tissue characterization.

METHODS

Standard-protocol (SP) and low-dose modified-protocol (MP) CTs were contemporaneously acquired in 34 Crohn's disease patients referred for CT. SP image reconstruction was via the manufacturer's recommendations (60% FBP, filtered back projection; 40% ASiR, Adaptive Statistical iterative Reconstruction; SP-ASiR40). MP data sets underwent four reconstructions (100% FBP; 40% ASiR; 70% ASiR; MBIR). Three observers measured tissue volumes using HU thresholds for fat, soft tissue and bone/contrast on each data set. Analysis was via SPSS.

RESULTS

Inter-observer agreement was strong for 1530 datapoints (rs > 0.9). MP-MBIR tissue volume measurement was superior to other MP reconstructions and closely correlated with the reference SP-ASiR40 images for all tissue types. MP-MBIR superiority was most marked for fat volume calculation - close SP-ASiR40 and MP-MBIR Bland-Altman plot correlation was seen with the lowest average difference (336 cm³ ) when compared with other MP reconstructions.

CONCLUSIONS

Hounsfield unit-determined tissue volume calculations from MP-MBIR images resulted in values comparable to SP-ASiR40 calculations and values that are superior to MP-ASiR images. Accuracy of estimation of volume of tissues (e.g. fat) using segmentation software on low-dose CT images appears optimal when reconstructed with pure IR.

Assessment of sub-milli-sievert abdominal computed tomography with iterative reconstruction techniques of different vendors

AIM: To assess diagnostic image quality of reduced dose (RD) abdominal computed tomography (CT) with 9 iterative reconstruction techniques (IRTs) from 4 different vendors to the standard of care (SD) CT.

METHODS: In an Institutional Review Board approved study, 66 patients (mean age 60 ± 13 years, 44 men, and 22 women) undergoing routine abdomen CT on multi-detector CT (MDCT) scanners from vendors A, B, and C (≥ 64 row CT scanners) (22 patients each) gave written informed consent for acquisition of an additional RD CT series. Sinogram data of RD CT was reconstructed with two vendor-specific and a vendor-neutral IRTs (A-1, A-2, A-3; B-1, B-2, B-3; and C-1, C-2, C-3) and SD CT series with filtered back projection. Subjective image evaluation was performed by two radiologists for each SD and RD CT series blinded and independently. All RD CT series (198) were assessed first followed by SD CT series (66). Objective image noise was measured for SD and RD CT series. Data were analyzed by Wilcoxon signed rank, kappa, and analysis of variance tests.

RESULTS: There were 13/50, 18/57 and 9/40 missed lesions (size 2-7 mm) on RD CT for vendor A, B, and C, respectively. Missed lesions includes liver cysts, kidney cysts and stone, gall stone, fatty liver, and pancreatitis. There were also 5, 4, and 4 pseudo lesions (size 2-3 mm) on RD CT for vendor A, B, and C, respectively. Lesions conspicuity was sufficient for clinical diagnostic performance for 6/24 (RD-A-1), 10/24 (RD-A-2), and 7/24 (RD-A-3) lesions for vendor A; 5/26 (RD-B-1), 6/26 (RD-B-2), and 7/26 (RD-B-3) lesions for vendor B; and 4/20 (RD-C-1) 6/20 (RD-C-2), and 10/20 (RD-C-3) lesions for vendor C (P = 0.9). Mean objective image noise in liver was significantly lower for RD A-1 compared to both RD A-2 and RD A-3 images (P < 0.001). Similarly, mean objective image noise lower for RD B-2 (compared to RD B-1, RD B-3) and RD C-3 (compared to RD C-1 and C-2) (P = 0.016).

CONCLUSION: Regardless of IRTs and MDCT vendors, abdominal CT acquired at mean CT dose index volume 1.3 mGy is not sufficient to retain clinical diagnostic performance.

Iterative Reconstruction Results in Larger Computed Tomography Measurements of Iliofemoral Artery Diameter in Patients Referred for Transcatheter Aortic Valve Replacement

OBJECTIVES

We hypothesized that improved iterative reconstruction increases image quality and reduces artifacts for iliofemoral artery computed tomography imaging in patients referred for transcatheter aortic valve replacement (TAVR).

METHODS

We examined 56 consecutive patients undergoing computed tomography for possible TAVR and compared image quality and iliofemoral artery size between adaptive statistical iterative reconstructions (ASIRs) and improved model-based iterative reconstructions (MBIRs).

RESULTS

Model-based iterative reconstruction (vs ASIR) was associated with improved (P < 0.001 for each) image quality (3.4 ± 0.8 vs 2.8 ± 1.0), beam hardening (3.5 ± 0.8 vs 3.0 ± 1.1), and wall definition (3.6 ± 0.6 vs 3.1 ± 0.8). Image signal-to-noise ratios (20.4 ± 10.1 vs 13.7 ± 6.6, P < 0.001) were also increased with MBIR as compared with ASIR. Mean iliofemoral artery size was larger using MBIR compared with ASIR (left, 7.7 ± 1.5 vs 7.4 ± 1.7 mm, P < 0.001; right, 7.8 ± 1.2 vs 7.4 ± 1.5 mm, P = 0.008).

CONCLUSIONS

In patients referred for TAVR, improved MBIR resulted in higher image quality, reduced artifacts, and larger iliofemoral artery diameters compared with standard iterative reconstructions.

Cumulative radiation exposure from diagnostic imaging in intensive care unit patients

AIM: To quantify cumulative effective dose of intensive care unit (ICU) patients attributable to diagnostic imaging.

METHODS: This was a prospective, interdisciplinary study conducted in the ICU of a large tertiary referral and level 1 trauma center. Demographic and clinical data including age, gender, date of ICU admission, primary reason for ICU admission, APACHE II score, length of stay, number of days intubated, date of death or discharge, and re-admission data was collected on all patients admitted over a 1-year period. The overall radiation exposure was quantified by the cumulative effective radiation dose (CED) in millisieverts (mSv) and calculated using reference effective doses published by the United Kingdom National Radiation Protection Board. Pediatric patients were selected for subgroup-analysis.

RESULTS: A total of 2737 studies were performed in 421 patients. The total CED was 1704 mSv with a median CED of 1.5 mSv (IQR 0.04-6.6 mSv). Total CED in pediatric patients was 74.6 mSv with a median CED of 0.07 mSv (IQR 0.01-4.7 mSv). Chest radiography was the most commonly performed examination accounting for 83% of all studies but only 2.7% of total CED. Computed tomography (CT) accounted for 16% of all studies performed and contributed 97% of total CED. Trauma patients received a statistically significant higher dose [median CED 7.7 mSv (IQR 3.5-13.8 mSv)] than medical [median CED 1.4 mSv (IQR 0.05-5.4 mSv)] and surgical [median CED 1.6 mSv (IQR 0.04-7.5 mSv)] patients. Length of stay in ICU [OR = 1.12 (95%CI: 1.079-1.157)] was identified as an independent predictor of receiving a CED greater than 15 mSv.

CONCLUSION: Trauma patients and patients with extended ICU admission times are at increased risk of higher CEDs. CED should be minimized where feasible, especially in young patients.

Effect of reconstruction methods and x-ray tube current-time product on nodule detection in an anthropomorphic thorax phantom: A crossed-modality JAFROC observer study

Purpose:

To evaluate nodule detection in an anthropomorphic chest phantom in computed tomography (CT) images reconstructed with adaptive iterative dose reduction 3D (AIDR^3D) and filtered back projection (FBP) over a range of tube current–time product (mAs).

Methods:

Two phantoms were used in this study: (i) an anthropomorphic chest phantom was loaded with spherical simulated nodules of 5, 8, 10, and 12 mm in diameter and +100, −630, and −800 Hounsfield units electron density; this would generate CT images for the observer study; (ii) a whole-body dosimetry verification phantom was used to ultimately estimate effective dose and risk according to the model of the BEIR VII committee. Both phantoms were scanned over a mAs range (10, 20, 30, and 40), while all other acquisition parameters remained constant. Images were reconstructed with both AIDR^3D and FBP. For the observer study, 34 normal cases (no nodules) and 34 abnormal cases (containing 1–3 nodules, mean 1.35 ± 0.54) were chosen. Eleven observers evaluated images from all mAs and reconstruction methods under the free-response paradigm. A crossed-modality jackknife alternative free-response operating characteristic (JAFROC) analysis method was developed for data analysis, averaging data over the two factors influencing nodule detection in this study: mAs and image reconstruction (AIDR^3D or FBP). A Bonferroni correction was applied and the threshold for declaring significance was set at 0.025 to maintain the overall probability of Type I error at α = 0.05. Contrast-to-noise (CNR) was also measured for all nodules and evaluated by a linear least squares analysis.

Results:

For random-reader fixed-case crossed-modality JAFROC analysis, there was no significant difference in nodule detection between AIDR^3D and FBP when data were averaged over mAs [F(1, 10) = 0.08, p = 0.789]. However, when data were averaged over reconstruction methods, a significant difference was seen between multiple pairs of mAs settings [F(3, 30) = 15.96, p < 0.001]. Measurements of effective dose and effective risk showed the expected linear dependence on mAs. Nodule CNR was statistically higher for simulated nodules on images reconstructed with AIDR^3D (p < 0.001).

Conclusions:

No significant difference in nodule detection performance was demonstrated between images reconstructed with FBP and AIDR^3D. mAs was found to influence nodule detection, though further work is required for dose optimization.

Knowledge-based iterative model reconstruction: comparative image quality and radiation dose with a pediatric computed tomography phantom

BACKGROUND

CT of pediatric phantoms can provide useful guidance to the optimization of knowledge-based iterative reconstruction CT.

OBJECTIVE

To compare radiation dose and image quality of CT images obtained at different radiation doses reconstructed with knowledge-based iterative reconstruction, hybrid iterative reconstruction and filtered back-projection.

MATERIALS AND METHODS

We scanned a 5-year anthropomorphic phantom at seven levels of radiation. We then reconstructed CT data with knowledge-based iterative reconstruction (iterative model reconstruction [IMR] levels 1, 2 and 3; Philips Healthcare, Andover, MA), hybrid iterative reconstruction (iDose(4), levels 3 and 7; Philips Healthcare, Andover, MA) and filtered back-projection. The noise, signal-to-noise ratio and contrast-to-noise ratio were calculated. We evaluated low-contrast resolutions and detectability by low-contrast targets and subjective and objective spatial resolutions by the line pairs and wire.

RESULTS

With radiation at 100 peak kVp and 100 mAs (3.64 mSv), the relative doses ranged from 5% (0.19 mSv) to 150% (5.46 mSv). Lower noise and higher signal-to-noise, contrast-to-noise and objective spatial resolution were generally achieved in ascending order of filtered back-projection, iDose(4) levels 3 and 7, and IMR levels 1, 2 and 3, at all radiation dose levels. Compared with filtered back-projection at 100% dose, similar noise levels were obtained on IMR level 2 images at 24% dose and iDose(4) level 3 images at 50% dose, respectively. Regarding low-contrast resolution, low-contrast detectability and objective spatial resolution, IMR level 2 images at 24% dose showed comparable image quality with filtered back-projection at 100% dose. Subjective spatial resolution was not greatly affected by reconstruction algorithm.

CONCLUSION

Reduced-dose IMR obtained at 0.92 mSv (24%) showed similar image quality to routine-dose filtered back-projection obtained at 3.64 mSv (100%), and half-dose iDose(4) obtained at 1.81 mSv.

Model-Based Iterative Reconstruction in CT Enterography

OBJECTIVE

The purpose of this study was to analyze the performance of pure model-based iterative reconstruction (MBIR) in low-dose CT enterography.

SUBJECTS AND METHODS

Forty-four patients with Crohn disease referred for CT enterography were included. Low-dose modified-protocol and conventional-protocol CT datasets were contemporaneously acquired. Conventional-protocol image formation was performed with 40% adaptive statistical iterative reconstruction (ASIR). Modified-protocol data were reconstructed with 100% MBIR and 40% ASIR. Image quality was assessed subjectively and objectively at six levels. Independent clinical interpretations by two fully blinded radiologists were compared with reference standard consensus reviews by two nonblinded readers who had access to clinical information, previous imaging studies, and medical records.

RESULTS

A 74.7% average radiation dose reduction was seen: low-dose modified-protocol effective dose, 1.61 ± 1.18 mSv (size-specific-dose-estimate, 2.47 ± 1.21 mGy); conventional-protocol effective dose, 6.05 ± 2.84 mSv (size-specific-dose-estimate, 9.25 ± 2.9 mGy). Image quality assessment yielded 9372 data points. Objective noise on modified-protocol MBIR images was superior (p < 0.05) to that with the conventional protocol at three of six levels and comparable at the other three levels. Modified-protocol images were superior to conventional-protocol ASIR images (p < 0.05 in all cases) for subjective noise, spatial resolution, contrast resolution, streak artifact, and diagnostic acceptability on coronal reconstructions. Axial diagnostic acceptability was superior for conventional-protocol ASIR (p = 0.76). For both readers, modified-protocol MBIR clinical readings agreed more closely with reference standard readings than did conventional-protocol ASIR readings with regard to bowel wall disease assessment (κ = 0.589 and 0.700 vs 0.583 and 0.564). Overall Crohn disease activity grade (κ = 0.549 and 0.441 vs 0.315 and 0.596) and detection of acute complications (κ = 1.0 and 0.689 vs 0.896 and 0.896) were comparable when evaluated on conventional-protocol ASIR and modified-protocol MBIR images.

CONCLUSION

Low-dose CT enterography with MBIR yields images that are comparable to or superior to conventional images.

Comparison of new and conventional versions of model-based iterative reconstruction in reduced-dose computed tomography for diagnosis of hepatic steatosis

PURPOSE

To compare new and conventional versions of model-based iterative reconstruction (MBIR) in reduced-dose computed tomography (CT) in terms of diagnostic performance for hepatic steatosis.

MATERIALS AND METHODS

Images were reconstructed from standard-dose and aggressively reduced-dose (the dose-length product was reduced by 91 %) unenhanced abdominopelvic CT scans of 86 patients using filtered back projection (SD-FBP) and new and conventional versions of MBIR (RD-MBIRn and RD-MBIRc), respectively. The mean CT attenuation of the liver (CT[L]) and the spleen as well as the ratio of these parameters (CT[L/S]) were calculated. CT[L] <48 Hounsfield units (HU) and CT[L/S] <1.1 were applied to SD-FBP (used as the reference standard; the number of positive patients was 12 and 14, respectively), RD-MBIRn, and RD-MBIRc.

RESULTS

CT[L]s in SD-FBP/RD-MBIRn/RD-MBIRc were 56.9/55.9/52.8 HU. The difference in CT[L] between RD-MBIRn and SD-FBP was within ±5.0 HU in most cases. The sensitivity/specificity/accuracy of CT[L] <48 HU in RD-MBIRn and RD-MBIRc were 1.00/0.97/0.98 and 1.00/0.92/0.93, respectively, showing that RD-MBIRn permits significant improvements in specificity and accuracy (P < 0.05, McNemar test). For CT[L/S] <1.1, these values were 0.79/0.97/0.94 and 0.79/0.97/0.94 in RD-MBIRn and RD-MBIRc, respectively.

CONCLUSION

When CT[L] <48 HU was applied, RD-MBIRn presented a significantly improved hepatic steatosis diagnostic performance compared with RD-MBIRc; indeed, it was almost equivalent to that afforded by SD-FBP.

Feasibility of low-dose CT with model-based iterative image reconstruction in follow-up of patients with testicular cancer

PURPOSE

We examine the performance of pure model-based iterative reconstruction with reduced-dose CT in follow-up of patients with early-stage testicular cancer.

METHODS

Sixteen patients (mean age 35.6 ± 7.4 years) with stage I or II testicular cancer underwent conventional dose (CD) and low-dose (LD) CT acquisition during CT surveillance. LD data was reconstructed with model-based iterative reconstruction (LD-MBIR). Datasets were objectively and subjectively analysed at 8 anatomical levels. Two blinded clinical reads were compared to gold-standard assessment for diagnostic accuracy.

RESULTS

Mean radiation dose reduction of 67.1% was recorded. Mean dose measurements for LD-MBIR were: thorax - 66 ± 11 mGy cm (DLP), 1.0 ± 0.2 mSv (ED), 2.0 ± 0.4 mGy (SSDE); abdominopelvic - 128 ± 38 mGy cm (DLP), 1.9 ± 0.6 mSv (ED), 3.0 ± 0.6 mGy (SSDE). Objective noise and signal-to-noise ratio values were comparable between the CD and LD-MBIR images. LD-MBIR images were superior (p < 0.001) with regard to subjective noise, streak artefact, 2-plane contrast resolution, 2-plane spatial resolution and diagnostic acceptability. All patients were correctly categorised as positive, indeterminate or negative for metastatic disease by 2 readers on LD-MBIR and CD datasets.

CONCLUSIONS

MBIR facilitated a 67% reduction in radiation dose whilst producing images that were comparable or superior to conventional dose studies without loss of diagnostic utility.

Model-based iterative reconstruction and adaptive statistical iterative reconstruction: dose-reduced CT for detecting pancreatic calcification

Background

Iterative reconstruction methods have attracted attention for reducing radiation doses in computed tomography (CT).

Purpose

To investigate the detectability of pancreatic calcification using dose-reduced CT reconstructed with model-based iterative construction (MBIR) and adaptive statistical iterative reconstruction (ASIR).

Material and Methods

This prospective study approved by Institutional Review Board included 85 patients (57 men, 28 women; mean age, 69.9 years; mean body weight, 61.2 kg). Unenhanced CT was performed three times with different radiation doses (reference-dose CT [RDCT], low-dose CT [LDCT], ultralow-dose CT [ULDCT]). From RDCT, LDCT, and ULDCT, images were reconstructed with filtered-back projection (R-FBP, used for establishing reference standard), ASIR (L-ASIR), and MBIR and ASIR (UL-MBIR and UL-ASIR), respectively. A lesion (pancreatic calcification) detection test was performed by two blinded radiologists with a five-point certainty level scale.

Results

Dose-length products of RDCT, LDCT, and ULDCT were 410, 97, and 36 mGy-cm, respectively. Nine patients had pancreatic calcification. The sensitivity for detecting pancreatic calcification with UL-MBIR was high (0.67–0.89) compared to L-ASIR or UL-ASIR (0.11–0.44), and a significant difference was seen between UL-MBIR and UL-ASIR for one reader (P = 0.014). The area under the receiver-operating characteristic curve for UL-MBIR (0.818–0.860) was comparable to that for L-ASIR (0.696–0.844). The specificity was lower with UL-MBIR (0.79–0.92) than with L-ASIR or UL-ASIR (0.96–0.99), and a significant difference was seen for one reader (P < 0.01).

Conclusion

In UL-MBIR, pancreatic calcification can be detected with high sensitivity, however, we should pay attention to the slightly lower specificity.

High-resolution CT with new model-based iterative reconstruction with resolution preference algorithm in evaluations of lung nodules: Comparison with conventional model-based iterative reconstruction and adaptive statistical iterative reconstruction

OBJECTIVE

To compare the image quality of high-resolution computed tomography (HRCT) for evaluating lung nodules reconstructed with the new version of model-based iterative reconstruction and spatial resolution preference algorithm (MBIRn) vs. conventional model-based iterative reconstruction (MBIRc) and adaptive statistical iterative reconstruction (ASIR).

MATERIALS AND METHODS

This retrospective clinical study was approved by our institutional review board and included 70 lung nodules in 58 patients (mean age, 71.2±10.9years; 34 men and 24 women). HRCT of lung nodules were reconstructed using MBIRn, MBIRc and ASIR. Objective image noise was measured by placing the regions of interest on lung parenchyma. Two blinded radiologists performed subjective image analyses.

RESULTS

Significant improvements in the following points were observed in MBIRn compared with ASIR (p<0.005): objective image noise (24.4±8.0 vs. 37.7±10.4), subjective image noise, streak artifacts, and adequateness for evaluating internal characteristics and borders of nodules. The sharpness of small vessels and bronchi and diagnostic acceptability with MBIRn were significantly better than with MBIRc and ASIR (p<0.008).

CONCLUSION

HRCT reconstructed with MBIRn provides diagnostically more acceptable images for the detailed analyses of lung nodules compared with MBIRc and ASIR.

Comparison of the image qualities of filtered back-projection, adaptive statistical iterative reconstruction, and model-based iterative reconstruction for CT venography at 80 kVp

PURPOSE

To evaluate the subjective and objective qualities of computed tomography (CT) venography images at 80 kVp using model-based iterative reconstruction (MBIR) and to compare these with those of filtered back projection (FBP) and adaptive statistical iterative reconstruction (ASIR) using the same CT data sets.

MATERIALS AND METHODS

Forty-four patients (mean age: 56.1 ± 18.1) who underwent 80 kVp CT venography (CTV) for the evaluation of deep vein thrombosis (DVT) during 4 months were enrolled in this retrospective study. The same raw data were reconstructed using FBP, ASIR, and MBIR. Objective and subjective image analysis were performed at the inferior vena cava (IVC), femoral vein, and popliteal vein.

RESULTS

The mean CNR of MBIR was significantly greater than those of FBP and ASIR and images reconstructed using MBIR had significantly lower objective image noise (p < .001). Subjective image quality and confidence of detecting DVT by MBIR group were significantly greater than those of FBP and ASIR (p < .005), and MBIR had the lowest score for subjective image noise (p < .001).

CONCLUSION

CTV at 80 kVp with MBIR was superior to FBP and ASIR regarding subjective and objective image qualities.

KEY POINTS

• MBIR provides superior image quality compared with FBP and ASIR • CTV at 80kVp with MBIR improves diagnostic confidence in diagnosing DVT • CTV at 80kVp with MBIR presents better image quality with low radiation.

CT of the Abdomen with Reduced Tube Voltage in Adults: A Practical Approach

Recent innovations in computed tomographic (CT) hardware and software have allowed implementation of low tube voltage imaging into everyday CT scanning protocols in adults. CT at a low tube voltage setting has many benefits, including (a) radiation dose reduction, which is crucial in young patients and those with chronic medical conditions undergoing serial CT examinations for disease management; and (b) higher contrast enhancement. For the latter, increased attenuation of iodinated contrast material improves the evaluation of hypervascular lesions, vascular structures, intestinal mucosa in patients with bowel disease, and CT urographic images. Additionally, the higher contrast enhancement may provide diagnostic images in patients with renal dysfunction receiving a reduced contrast material load and in patients with suboptimal peripheral intravenous access who require a lower contrast material injection rate. One limitation is that noisier images affect image quality at a low tube voltage setting. The development of denoising algorithms such as iterative reconstruction has made it possible to perform CT at a low tube voltage setting without compromising diagnostic confidence. Other potential pitfalls of low tube voltage CT include (a) photon starvation artifact in larger patients, (b) accentuation of streak artifacts, and (c) alteration of the CT attenuation value, which may affect evaluation of lesions on the basis of conventional enhancement thresholds. CT of the abdomen with a low tube voltage setting is an excellent radiation reduction technique when properly applied to imaging of select patients in the appropriate clinical setting.

The role of pure iterative reconstruction in conventional dose CT enterography

PURPOSE

Pure iterative reconstruction (Pure IR) has been proposed as a solution to improve diagnostic quality of low dose CT images. We assess the performance of model based iterative reconstruction (MBIR) in improving conventional dose CT enterography (CTE) images.

METHODS

43 Crohn's patients (27 female) (38.5 ± 12.98 years) referred for CTE were included. Images were reconstructed with pure IR (MBIR, General Electric Healthcare) in addition to standard department protocol (reconstructed with hybrid iterative reconstruction (Hybrid IR) [60% filtered back projection/40% adaptive statistical IR (General Electric Healthcare)]. Image quality was assessed objectively and subjectively at 6 anatomical levels. Clinical interpretation was undertaken in consensus by 2 blinded radiologists along with 2 non-blinded readers ('gold standard'). Results were analyzed using Statistical Package for Social Scientists.

RESULTS

Mean effective radiation dose was 6.05 ± 2.84 mSv (size specific dose estimates 9.25 ± 2.9 mGy). Objective and subjective assessment yielded 6106 data points. Pure IR images significantly outperformed those using standard reconstruction techniques across all subjective (p < 0.001 for all comparisons) (noise, contrast resolution, spatial resolution, streak artifact, axial diagnostic acceptability, coronal diagnostic acceptability) and objective (p < 0.004) (noise, signal-to-noise ratio) parameters. Clinical reads of the pure IR images agreed more closely with the gold standard reads than the hybrid IR image reads in terms of overall Crohn's activity grade (κ = 0.630, 0.308) and detection of acute complications (κ = 1.0, 0.896). Results were comparable for bowel wall disease severity assessment (κ = 0.523, 0.593).

CONCLUSIONS

Pure IR considerably improves image quality of conventional dose CTE images and therefore its use should be expanded beyond low dose protocols to improving image quality at conventional dose CT imaging.

Assessment of Filtered Back Projection, Adaptive Statistical, and Model-Based Iterative Reconstruction for Reduced Dose Abdominal Computed Tomography

PURPOSE

To compare standard of care and reduced dose (RD) abdominal computed tomography (CT) images reconstructed with filtered back projection (FBP), adaptive statistical iterative reconstruction (ASIR), model-based iterative reconstruction (MBIR) techniques.

MATERIALS AND METHODS

In an Institutional Review Board-approved, prospective clinical study, 28 patients (mean age 59 ± 13 years ), undergoing clinically indicated routine abdominal CT on a 64-channel multi-detector CT scanner, gave written informed consent for acquisition of an additional RD (<1 milli-Sievert) abdomen CT series. Sinogram data of RD series were reconstructed with FBP, ASIR, and MBIR and compared with FBP images of standard dose abdomen CT. Two radiologists performed randomized, independent, and blinded comparison for lesion detection, lesion margin, visibility of normal structures, and diagnostic confidence.

RESULTS

Mean CT dose index volume was 10 ± 3.4 mGy and 1.3 ± 0.3 mGy for standard and RD CT, respectively. There were 73 "true positive" lesions detected on standard of care CT. Nine lesions (<8 mm in size) were missed on RD abdominal CT images which included liver lesions, liver cysts, kidney cysts, and paracolonic abscess. These lesions were missed regardless of patient size and types of iterative reconstruction techniques used for reconstruction of RD data sets. The visibility of lesion margin was suboptimal in (23/28) patients with RD FBP, (15/28) patients with RD ASIR, and (14/28) patients with RD MBIR compared to standard of care FBP images (P < 0.001). Diagnostic confidence for the assessment of lesions on RD images was suboptimal in most patients regardless of iterative reconstruction techniques.

CONCLUSIONS

Clinically significant lesions (< 8 mm) can be missed on abdominal CT examinations acquired at a CT dose index volume of 1.3 mGy regardless of patients' size and reconstruction techniques (FBP, ASIR, and MBIR).

Improving head and neck CTA with hybrid and model-based iterative reconstruction techniques

AIM

To compare image quality of head and neck computed tomography angiography (CTA) reconstructed with filtered back projection (FBP), hybrid iterative reconstruction (HIR) and model-based iterative reconstruction (MIR) algorithms.

MATERIALS AND METHODS

The raw data of 34 studies were simultaneously reconstructed with FBP, HIR (iDose(4), Philips Healthcare, Best, the Netherlands), and with a prototype version of a MIR algorithm (IMR, Philips Healthcare). Objective (contrast-to-noise ratio [CNR], vascular contrast, automatic vessel analysis [AVA], stenosis grade) and subjective image quality (ranking at level of the circle of Willis, carotid bifurcation, and shoulder) of the five reconstructions were compared using repeated-measures analysis of variance (ANOVA) and post-hoc analysis.

RESULTS

Vascular contrast was significantly higher in both the circle of Willis and carotid bifurcation with both levels of MIR compared to the other reconstruction methods (all p<0.0001). The CNR was highest for high MIR, followed by low MIR, high HIR, mid HIR and FBP (p<0.001 except low MIR versus high HIR; p>0.33). AVA showed most complete carotids in both MIR-levels, followed by high HIR (p>0.08), mid HIR (p<0.023) and FBP (p<0.010), vertebral arteries completeness was similar (p=0.40 and p=0.06). Stenosis grade showed no significant differences (p=0.16). High HIR showed the best subjective image quality at the circle of Willis and carotid bifurcation level, followed by mid HIR. At shoulder level, low MIR and high HIR were ranked best, followed by high MIR.

CONCLUSION

Objectively, MIR significantly improved the overall image quality, reduced image noise, and improved automated vessel analysis, whereas FBP showed the lowest objective image quality. Subjectively, the highest level of HIR was considered superior at the level of the circle of Willis and the carotid bifurcation, and along with the lowest level of MIR for the origins of the neck arteries at shoulder level.

Model-based iterative reconstruction in low-dose CT colonography-feasibility study in 65 patients for symptomatic investigation

RATIONALE AND OBJECTIVES

To compare image quality on computed tomographic colonography (CTC) acquired at standard dose (STD) and low dose (LD) using filtered-back projection, adaptive statistical iterative reconstruction, and model-based iterative reconstruction (MBIR) techniques.

MATERIALS AND METHODS

A total of 65 symptomatic patients were prospectively enrolled for the study and underwent STD and LD CTC with filtered-back projection, adaptive statistical iterative reconstruction, and MBIR to allow direct per-patient comparison. Objective image noise, subjective image analyses, and polyp detection were assessed.

RESULTS

Objective image noise analysis demonstrates significant noise reduction using MBIR technique (P < .05) despite being acquired at lower doses. Subjective image analyses were superior for LD MBIR in all parameters except visibility of extracolonic lesions (two-dimensional) and visibility of colonic wall (three-dimensional) where there were no significant differences. There was no significant difference in polyp detection rates (P > .05). Doses: LD (dose-length product, 257.7), STD (dose-length product, 483.6).

CONCLUSIONS

LD MBIR CTC objectively shows improved image noise using parameters in our study. Subjectively, image quality is maintained. Polyp detection shows no significant difference but because of small numbers needs further validation. Average dose reduction of 47% can be achieved. This study confirms feasibility of using MBIR in this context of CTC in symptomatic population.

CT of the pancreas: comparison of image quality and pancreatic duct depiction among model-based iterative, adaptive statistical iterative, and filtered back projection reconstruction techniques

The purpose of this study is to compare CT images of the pancreas reconstructed with model-based iterative reconstruction (MBIR), adaptive statistical iterative reconstruction (ASiR), and filtered back projection (FBP) techniques for image quality and pancreatic duct (PD) depiction. Data from 40 patients with contrast-enhanced abdominal CT [CTDIvol: 10.3 ± 3.0 (mGy)] during the late arterial phase were reconstructed with FBP, 40% ASiR-FBP blending, and MBIR. Two radiologists assessed the depiction of the main PD, image noise, and overall image quality using 5-point scale independently. Objective CT value and noise were measured in the pancreatic parenchyma, and the contrast-to-noise ratio (CNR) of the PD was calculated. The Friedman test and post-hoc multiple comparisons with Bonferroni test following one-way ANOVA were used for qualitative and quantitative assessment, respectively. For the subjective assessment, scores for MBIR were significantly higher than those for FBP and 40% ASiR (all P < 0.001). No significant differences in CT values of the pancreatic parenchyma were noted among FBP, 40% ASiR, and MBIR images (P > 0.05). Objective image noise was significantly lower and CNR of the PD was higher with MBIR than with FBP and 40% ASiR (all P < 0.05). Our results suggest that pancreatic CT images reconstructed with MBIR have lower image noise, better image quality, and higher conspicuity and CNR of the PD compared with FBP and ASiR.

Impact of model-based iterative reconstruction on image quality of contrast-enhanced neck CT

BACKGROUND AND PURPOSE

Improved image quality is clinically desired for contrast-enhanced CT of the neck. We compared 30% adaptive statistical iterative reconstruction and model-based iterative reconstruction algorithms for the assessment of image quality of contrast-enhanced CT of the neck.

MATERIALS AND METHODS

Neck contrast-enhanced CT data from 64 consecutive patients were reconstructed retrospectively by using 30% adaptive statistical iterative reconstruction and model-based iterative reconstruction. Objective image quality was assessed by comparing SNR, contrast-to-noise ratio, and background noise at levels 1 (mandible) and 2 (superior mediastinum). Two independent blinded readers subjectively graded the image quality on a scale of 1-5, (grade 5 = excellent image quality without artifacts and grade 1 = nondiagnostic image quality with significant artifacts). The percentage of agreement and disagreement between the 2 readers was assessed.

RESULTS

Compared with 30% adaptive statistical iterative reconstruction, model-based iterative reconstruction significantly improved the SNR and contrast-to-noise ratio at levels 1 and 2. Model-based iterative reconstruction also decreased background noise at level 1 (P = .016), though there was no difference at level 2 (P = .61). Model-based iterative reconstruction was scored higher than 30% adaptive statistical iterative reconstruction by both reviewers at the nasopharynx (P < .001) and oropharynx (P < .001) and for overall image quality (P < .001) and was scored lower at the vocal cords (P < .001) and sternoclavicular junction (P < .001), due to artifacts related to thyroid shielding that were specific for model-based iterative reconstruction.

CONCLUSIONS

Model-based iterative reconstruction offers improved subjective and objective image quality as evidenced by a higher SNR and contrast-to-noise ratio and lower background noise within the same dataset for contrast-enhanced neck CT. Model-based iterative reconstruction has the potential to reduce the radiation dose while maintaining the image quality, with a minor downside being prominent artifacts related to thyroid shield use on model-based iterative reconstruction.

Model-based iterative reconstruction for improvement of low-contrast detectability in liver CT at reduced radiation dose: ex-vivo experience

AIM

To compare low-contrast detectability, and qualitative and quantitative image parameters on standard and reduced radiation dose abdominal CT reconstructed with filtered back projection (FBP) and model-based iterative reconstruction (MBIR).

MATERIALS AND METHODS

A custom built liver phantom containing 43 lesions was imaged at 120 kVp and four radiation dose levels (100% = 188 mAs, 50%, 25%, and 10%). Image noise and contrast-to-noise ratios (CNR) were assessed. Lesion detection and qualitative image analysis (five-point Likert scale with 1 = worst, 5 = best for confidence) was performed by three independent radiologists.

RESULTS

CNR on MBIR images was significantly higher (mean 246%, range 151-383%) and image noise was significantly lower (69%, 59-78%) than on FBP images at the same radiation dose (both p < 0.05). On MBIR 10% images, CNR (3.3 ± 0.3) was significantly higher and noise (15 ± 1HU) significantly lower than on FBP 100% images (2.5 ± 0.1; 21 ± 1 HU). On 100% images, lesion attenuation was significantly lower with MBIR than with FBP (mean difference -2 HU). Low-contrast detectability and qualitative results were similar with MBIR 50% and FBP 100%.

CONCLUSION

Low-contrast detectability with MBIR 50% and FBP 100% were equal. Quantitative parameters on even lower dose MBIR images are superior to 100%-dose FBP images. Some attenuation values differ significantly with MBIR compared with FBP.

Statistical model based iterative reconstruction (MBIR) in clinical CT systems: experimental assessment of noise performance

PURPOSE

To reduce radiation dose in CT imaging, the statistical model based iterative reconstruction (MBIR) method has been introduced for clinical use. Based on the principle of MBIR and its nonlinear nature, the noise performance of MBIR is expected to be different from that of the well-understood filtered backprojection (FBP) reconstruction method. The purpose of this work is to experimentally assess the unique noise characteristics of MBIR using a state-of-the-art clinical CT system.

METHODS

Three physical phantoms, including a water cylinder and two pediatric head phantoms, were scanned in axial scanning mode using a 64-slice CT scanner (Discovery CT750 HD, GE Healthcare, Waukesha, WI) at seven different mAs levels (5, 12.5, 25, 50, 100, 200, 300). At each mAs level, each phantom was repeatedly scanned 50 times to generate an image ensemble for noise analysis. Both the FBP method with a standard kernel and the MBIR method (Veo(®), GE Healthcare, Waukesha, WI) were used for CT image reconstruction. Three-dimensional (3D) noise power spectrum (NPS), two-dimensional (2D) NPS, and zero-dimensional NPS (noise variance) were assessed both globally and locally. Noise magnitude, noise spatial correlation, noise spatial uniformity and their dose dependence were examined for the two reconstruction methods.

RESULTS

(1) At each dose level and at each frequency, the magnitude of the NPS of MBIR was smaller than that of FBP. (2) While the shape of the NPS of FBP was dose-independent, the shape of the NPS of MBIR was strongly dose-dependent; lower dose lead to a "redder" NPS with a lower mean frequency value. (3) The noise standard deviation (σ) of MBIR and dose were found to be related through a power law of σ ∝ (dose)(-β) with the component β ≈ 0.25, which violated the classical σ ∝ (dose)(-0.5) power law in FBP. (4) With MBIR, noise reduction was most prominent for thin image slices. (5) MBIR lead to better noise spatial uniformity when compared with FBP. (6) A composite image generated from two MBIR images acquired at two different dose levels (D1 and D2) demonstrated lower noise than that of an image acquired at a dose level of D1+D2.

CONCLUSIONS

The noise characteristics of the MBIR method are significantly different from those of the FBP method. The well known tradeoff relationship between CT image noise and radiation dose has been modified by MBIR to establish a more gradual dependence of noise on dose. Additionally, some other CT noise properties that had been well understood based on the linear system theory have also been altered by MBIR. Clinical CT scan protocols that had been optimized based on the classical CT noise properties need to be carefully re-evaluated for systems equipped with MBIR in order to maximize the method's potential clinical benefits in dose reduction and/or in CT image quality improvement.

Image comparative assessment using iterative reconstructions: clinical comparison of low-dose abdominal/pelvic computed tomography between adaptive statistical, model-based iterative reconstructions and traditional filtered back projection in 65 patients

OBJECTIVES

The objective of this study was to compare image quality (objective and subjective parameters) and confidence in lesion detection between 3 image reconstruction algorithms in computed tomographic (CT) examinations of the abdomen/pelvis.

MATERIALS AND METHODS

This prospective institutional review board-approved study included 65 patients (mean [SD] age, 71.3 ± 9 years; mean [SD] body mass index, 24.4 [4.8] kg) who underwent routine CT examinations of the abdomen/pelvis followed immediately by 2 low-dose scans. Raw data sets were reconstructed by using filtered back projection (FBP), adaptive statistical iterative reconstruction (ASIR), and a model-based iterative reconstruction (MBIR). Measurements of objective noise and CT numbers were compared using repeated-measures analysis of variance. Six subjective image quality parameters were scored. Diagnostic confidence and accuracy in detection of various elementary lesions were performed.

RESULTS

Objectively, mean image noise for MBIR was significantly superior at all dose levels (P < 0.001). Subjectively, standard-dose ASIR and low-dose MBIR scans were better than standard-dose FBP scan in all parameters assessed (P < 0.05). Low-dose MBIR scans were comparable with standard-dose ASIR scans in all parameters except at noise index of 70 (approximately 85% dose reduction), where, in this case, the detection of liver lesions less than 5 mm were rated inferior (P < 0.05) with diagnostic accuracy reducing to 77.4%.

CONCLUSIONS

Low-dose MBIR scan shows superior objective noise reduction compared with standard-dose FBP and ASIR. Subjectively, low-dose MBIR scans at 76% dose reduction were also superior compared with standard-dose FBP and ASIR. However, at dose reductions of 85%, small liver lesions may be missed.

Advanced modelled iterative reconstruction for abdominal CT: qualitative and quantitative evaluation

AIM

To determine qualitative and quantitative image-quality parameters in abdominal imaging using advanced modelled iterative reconstruction (ADMIRE) with third-generation dual-source 192 section CT.

MATERIALS AND METHODS

Forty patients undergoing abdominal portal-venous CT at different tube voltage levels (90, 100, 110, and 120 kVp, n = 10 each) and 10 consecutive patients undergoing abdominal non-enhanced low-dose CT (100 kVp, 60 mAs) using a third-generation dual-source 192 section CT machine in the single-source mode were included. Images were reconstructed with filtered back projection (FBP) and ADMIRE (strength levels 1-5). Two blinded, independent readers subjectively determined image noise, artefacts, visibility of small structures, and image contrast, and measured attenuation in the liver, spleen, kidney, muscle, fat, and urinary bladder, and objective image noise.

RESULTS

Subjective noise was significantly lower and image contrast significantly higher for each increasing ADMIRE strength level and also for ADMIRE 1 compared to FBP (all, p < 0.001). No significant differences were found for artefact and visibility ratings among image sets (all, p > 0.05). Attenuation was similar across tube voltage-image datasets in all anatomical regions (all, p > 0.05). Objective noise was significantly lower for each increasing ADMIRE strength level, and for ADMIRE 1 compared to FBP (all, p < 0.001, maximal reduction 53%). Independent predictors of noise were tube voltage (p < 0.05) and current (p < 0.001), diameter (p < 0.05), and reconstruction algorithm (p<0.001); the amount of noise reduction was related only to the reconstruction algorithm (p < 0.001).

CONCLUSION

Abdominal CT using ADMIRE results in an improved image quality with lower image noise as compared with FBP, while the attenuation of various anatomical regions remains constant among reconstruction algorithms.

Choosing the best reconstruction technique in abdominal computed tomography: a systematic approach

OBJECTIVE

There is uncertainty regarding the effect of iterative reconstruction (IR) techniques and other reconstruction algorithms on image quality. The aim of this study was to optimize image quality in relation to radiation dose in computed tomography (CT) liver examinations by comparing images reconstructed with different abdominal filters with and without IR.

METHODS

An anthropomorphic phantom was scanned on a Toshiba Aquilion ONE CT scanner. Images at 2 different dose levels were reconstructed with 12 different body reconstruction filters, all with both filtered back-projection and Adaptive Iterative Dose Reduction 3 dimensional. Receiver operating characteristic curves were constructed. The 2 reconstruction combinations with the highest scores from the phantom study were evaluated in a second comparison of clinical images. Six liver examinations were reconstructed with both filters and evaluated using visual grading analysis.

RESULTS

Two combinations of reconstruction filters and IR were the only 2 options among the 8 best images at both dose levels (area under the curve, 0.96 and 0.94 for 15 mGy as well as 0.86 and 0.84 for 10 mGy). In the patient study, one of these filters in combination with IR scored slightly higher than the other in combination with IR (mean score, 2.60 and 2.57, respectively; P = 0.56). Iterative reconstruction did not significantly increase lesion detectability for any of the filters.

CONCLUSIONS

This study indicates that the preferred choice for reconstruction of CT liver examinations performed with the Toshiba Aquilion ONE should be the FC18 filter with IR, although the IR technique did not significantly improve lesion detectability and did not compensate for the dose reduction in this study.