Deep learning image reconstruction for quality assessment of iodine concentration in computed tomography: A phantom study

BACKGROUND

Recently, deep learning reconstruction (DLR) technology aiming to improve image quality with minimal radiation dose has been applied not only to pediatric scans, but also to computed tomography angiography (CTA).

OBJECTIVE

To evaluate image quality characteristics of filtered back projection (FBP), hybrid iterative reconstruction [Adaptive Iterative Dose Reduction 3D (AIDR 3D)], and DLR (AiCE) using different iodine concentrations and scan parameters.

METHODS

Phantoms with eight iodine concentrations (ranging from 1.2 to 25.9 mg/mL) located at the edge of a cylindrical water phantom with a diameter of 19 cm were scanned. Data were reconstructed with FBP, AIDR 3D, and AiCE using various scan parameters of tube current and voltage using a 320 row-detector CT scanner. Data obtained using different reconstruction techniques were quantitatively compared by analyzing Hounsfield units (HU), noise, and contrast-to-noise ratios (CNRs).

RESULTS

HU values of FBP and AIDR 3D were constant even when the iodine concentration was changed, whereas AiCE showed the highest HU value when the iodine concentration was low, but the HU value reversed when the iodine concentration exceeded a certain value. In the AIDR 3D and AiCE, the noise decreased as the tube current increased, and the change in noise when the iodine concentration was inconsistent. AIDR 3D and AiCE yielded better noise reduction rates than with FBP at a low tube current. The noise reduction rate of AIDR 3D and AiCE compared to that of FBP showed characteristics ranging from 7% to 35%, and the noise reduction rate of AiCE compared to that of AIDR 3D ranged from 2.0% to 13.3%.

CONCLUSIONS

The evaluated reconstruction techniques showed different image quality characteristics (HU value, noise, and CNR) according to dose and scan parameters, and users must consider these results and characteristics before performing patient scans.

Iterative Reconstruction in Dose Reduction of A Head CT Examination and Corresponding Acquisition Parameter Selection

PURPOSE

To investigate the potential of iterative reconstruction in radiation dose reduction during head computed tomography (CT) examinations and to evaluate the relationship between the parameters milliampere second (mAs), kilovoltage (kV), and iterative reconstruction strength using a live ovine (sheep) model.

METHODS

A sheep was scanned on a SOMATOM Force (Siemens Healthineers) CT scanner at 12 mAs and 3 kV. Images were reconstructed with filtered back projection (FBP) and the Advanced Modeled Iterative Reconstruction (ADMIRE; Siemens Healthineers) strengths 1 to 5. Images with 216 combinations of varying doses, kVs, and reconstructions were rated by 2 neuroradiologists for low-contrast detectability (ie, gray-white matter differentiation) and image texture.

RESULTS

Using only gray-white matter differentiation, maximum dose reduction was 75% at 100 kV with ADMIRE-3, and using only image texture, maximum dose reduction was 75% at 120 kV (and 140 kV) with ADMIRE-5. When these 2 metrics were combined, maximum dose reduction was 50% at 120 kV with ADMIRE-3. Other kV levels and higher iterative reconstruction strengths did not offer superior results.

DISCUSSION

Although artificial intelligence algorithms are certainly gaining momentum, iterative reconstruction technology likely will remain more accessible to most hospitals and imaging centers. Dose reduction with preservation of image quality (ie, gray-white differentiation and image texture) can be achieved when complemented by appropriate iterative reconstruction strength. However, the effect of iterative reconstruction strength on gray-white differentiation and image texture does not necessarily converge on the same pattern.

CONCLUSION

Maximum dose reduction was 50% at 120 kV with ADMIRE-3, which confirms the potential for dose reduction with appropriately chosen iterative reconstruction strength and reveals a preference for 120 kV, as well as a limit to dose reduction by further increasing iterative reconstruction strength. A better understanding of dose-voltage-reconstruction relationships in iterative reconstruction might allow for greater dose reductions than current practices allow.

Influence of Adaptive Statistical Iterative Reconstructions on CT Radiomic Features in Oncologic Patients

Iterative reconstructions (IR) might alter radiomic features extraction. We aim to evaluate the influence of Adaptive Statistical Iterative Reconstruction-V (ASIR-V) on CT radiomic features. Patients who underwent unenhanced abdominal CT (Revolution Evo, GE Healthcare, USA) were retrospectively enrolled. Raw data of filtered-back projection (FBP) were reconstructed with 10 levels of ASIR-V (10–100%). CT texture analysis (CTTA) of liver, kidney, spleen and paravertebral muscle for all datasets was performed. Six radiomic features (mean intensity, standard deviation (SD), entropy, mean of positive pixel (MPP), skewness, kurtosis) were extracted and compared between FBP and all ASIR-V levels, with and without altering the spatial scale filter (SSF). CTTA of all organs revealed significant differences between FBP and all ASIR-V reconstructions for mean intensity, SD, entropy and MPP (all p < 0.0001), while no significant differences were observed for skewness and kurtosis between FBP and all ASIR-V reconstructions (all p > 0.05). A per-filter analysis was also performed comparing FBP with all ASIR-V reconstructions for all six SSF separately (SSF0-SSF6). Results showed significant differences between FBP and all ASIR-V reconstruction levels for mean intensity, SD, and MPP (all filters p < 0.0315). Skewness and kurtosis showed no differences for all comparisons performed (all p > 0.05). The application of incremental ASIR-V levels affects CTTA across various filters. Skewness and kurtosis are not affected by IR and may be reliable quantitative parameters for radiomic analysis.

[A Simple Method for Computationally Generating Metal Artifacts in CT Images for Treatment Planning: A Pilot Phantom Study]

PURPOSE

In treatment planning for radiation therapy, the use of computed tomography (CT) images including metal artifacts causes a reduction in the dose calculation accuracy. In clinical practice, the artifacts are manually contoured and assigned an appropriate fixed CT number. To validate the procedure, images taken before and after metal insertion into a patient are required, which may be impractical. We propose a simple method for computationally generating metal artifacts in clinical images.

METHODS

In the proposed method, a clinical image free of metal artifacts is used. To simulate metal inside a patient, CT numbers of a region in the image are replaced with a fixed extremely high value. A sinogram is created by the forward projection of the image. Data values of the sinogram in the metal region are converted into smaller values. From the sinogram, an image including artifacts is reconstructed with the filtered back projection.

RESULTS

The simulated artifacts consisted of dark and bright bands and were observed to be similar to the actual metal artifacts. CT numbers in multiple small regions of interest in the image obtained by the proposed method showed a good agreement with those in the actual image.

CONCLUSION

The proposed method was demonstrated to generate the metal artifacts additionally on the clinical images. The method would be potentially applicable to a validation study for the clinical procedure of manually contouring and assigning CT numbers to metal artifacts.

Impact of a New Adaptive Statistical Iterative Reconstruction (ASIR)-V Algorithm on Image Quality in Coronary Computed Tomography Angiography

RATIONALE AND OBJECTIVES

A new postprocessing algorithm named adaptive statistical iterative reconstruction (ASIR)-V has been recently introduced. The aim of this article was to analyze the impact of ASIR-V algorithm on signal, noise, and image quality of coronary computed tomography angiography.

MATERIALS AND METHODS

Fifty consecutive patients underwent clinically indicated coronary computed tomography angiography (Revolution CT; GE Healthcare, Milwaukee, WI). Images were reconstructed using filtered back projection and ASIR-V 0%, and a combination of filtered back projection and ASIR-V 20%-80% and ASIR-V 100%. Image noise, signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR) were calculated for left main coronary artery (LM), left anterior descending artery (LAD), left circumflex artery (LCX), and right coronary artery (RCA) and were compared between the different postprocessing algorithms used. Similarly a four-point Likert image quality score of coronary segments was graded for each dataset and compared. A cutoff value of P < .05 was considered statistically significant.

RESULTS

Compared to ASIR-V 0%, ASIR-V 100% demonstrated a significant reduction of image noise in all coronaries (P < .01). Compared to ASIR-V 0%, SNR was significantly higher with ASIR-V 60% in LM (P < .01), LAD (P < .05), LCX (P < .05), and RCA (P < .01). Compared to ASIR-V 0%, CNR for ASIR-V ≥60% was significantly improved in LM (P < .01), LAD (P < .05), and RCA (P < .01), whereas LCX demonstrated a significant improvement with ASIR-V ≥80%. ASIR-V 60% had significantly better Likert image quality scores compared to ASIR-V 0% in segment-, vessel-, and patient-based analyses (P < .01).

CONCLUSIONS

Reconstruction with ASIR-V 60% provides the optimal balance between image noise, SNR, CNR, and image quality.

Comparison of Full- and Half-Dose Image Reconstruction With Filtered Back Projection or Sinogram-Affirmed Iterative Reconstruction in Dual-Source Single-Energy MDCT Urography

OBJECTIVE

The purpose of this study is to prospectively compare the image quality of and confidence in the presence of a lesion on CT urography images acquired using filtered back projection (FBP) with 100% and 50% radiation doses with those for images simultaneously acquired using sinogram-affirmed iterative reconstruction with strength 3 (SAFIRE) with 50% and 25% radiation doses for patients with a high risk for urothelial carcinomas.

SUBJECTS AND METHODS

A total of 150 patients randomly underwent CT urography examinations performed using a dual-source single-energy scanner. After the radiation output of each tube was adjusted, datasets at three radiation dose levels were reconstructed using FBP and SAFIRE. Seven radiologists subjectively assessed image quality and confidence in the presence of a lesion for a total of 1200 datasets. Nonparametric methods for cluster data were used to estimate AUC values for variance methods on the basis of a noninferiority margin of 0.05.

RESULTS

The mean AUC value for image quality in SAFIRE with a 25% radiation dose was significantly lower than that of FBP with 100% radiation dose (p < 0.05 for all). The mean AUC values for the presence of a lesion were 0.907 and 0.894 for FBP, respectively, at 100% and 50% radiation doses, respectively, and 0.900 and 0.799 for SAFIRE at 50% and 25% radiation doses, respectively. However, the image quality of images acquired with SAFIRE at a 25% radiation dose was significantly inferior to that of images acquired with FBP at a 100% radiation dose.

CONCLUSION

Regardless of the experience of the radiologist, CT urography images acquired with FBP and SAFIRE with a 50% radiation dose were noninferior to those acquired with FBP with a 100% radiation dose in terms of image quality and confidence in the presence of a lesion, whereas those acquired with SAFIRE with 25% radiation dose were inferior.

Knowledge-based iterative model reconstruction: Comparative image quality with low tube voltage cerebral CT angiography

The aim of this study was to compare image quality of low tube voltage cerebral computed tomography angiography (CTA) reconstructed with knowledge-based iterative model reconstruction (IMR), filtered back projection (FBP), and hybrid iterative reconstruction (HIR).A total of 101 patients with suspected cerebrovascular diseases were enrolled and randomized into 2 groups, 100 kVp tube voltage (n = 53) and reduced tube voltage (80 kVp) (n = 48). Computed tomography data were reconstructed with IMR, FBP, and HIR algorithms. The image noise, vascular attenuation, signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR) were measured and calculated. Two blinded radiologists independently evaluated image quality based on diagnostic confidence on a 3-point scale. Quantitative and qualitative assessments were compared between different groups and reconstruction subgroups.Vascular attenuation was higher in the reduced tube voltage group than in 100-kVp tube voltage group, but showed no significant difference within each group. In both groups, the image noise, vascular SNR, and CNR were significantly improved by IMR as compared with FBP and HIR. Inter-group comparison indicated that IMR with reduced tube voltage showed better image quality with lower image noise and higher vascular SNR and CNR than FBP and HIR at 100 kVp, but slightly inferior to IMR at 100 kVp. IMR also yields the best qualitative image quality, and improves the diagnostic confidence of atherosclerosis and aneurysm. Compared with the standard 120-kVp protocol (1.86mSv), the radiation doses of 100 kVp (1.13mSv) and 80 kVp (0.56mSv) were 39% and 70% less, respectively.The quantitative and qualitative image quality obtained by IMR was superior to that obtained by FBP and HIR for low tube voltage cerebral CTA.

Correction Factors for CT Coronary Artery Calcium Scoring Using Advanced Modeled Iterative Reconstruction Instead of Filtered Back Projection

RATIONALE AND OBJECTIVES

Iterative reconstruction (IR) computed tomography (CT) techniques allow for radiation dose reduction while maintaining image quality. However, CT coronary artery calcium (CAC) scores may be influenced by certain IR algorithms. The aim of our study is to identify suitable correction factors to ensure consistency between IR and filtered back projection (FBP)-based CAC scoring.

MATERIAL AND METHODS

A phantom study was performed to derive suitable correction factors for CAC scores and volume (VOL) values with advanced modeled iterative reconstruction (or ADMIRE) strength level 3 (ADM3) and 5 (ADM5) vs FBP. CT data from 40 patients were retrospectively analyzed, and CAC score and VOL values were obtained following reconstruction with FBP, ADM3, and ADM5. Linear regression analysis was performed to obtain correction factors. Results with and without application of the correction factors were compared. Inter-reader agreement for risk class stratification was analyzed.

RESULTS

Phantom experiments determined a correction factor of 1.14 for ADM3 and 1.25 for ADM5. FBP-based CAC scores (897 ± 1413) were significantly higher than uncorrected scores with ADM3 (746 ± 1184, P ≤ .001) and ADM5 (640 ± 1036, P ≤ .001). After application of correction factors, no significant differences were found for CAC scores based on FBP (897 ± 1413) and ADM3 (853 ± 1353, P = .07). The inter-reader agreement for risk stratification was excellent (k = 0.91).

CONCLUSION

ADM3 can be applied to CAC scoring with use of a correction factor. When applying a correction factor of 1.14, excellent agreement with standard FBP for both CAC score and VOL can be achieved.

Radiation dose and diagnostic image quality associated with iterative reconstruction in coronary CT angiography: A systematic review

The aim of this systematic review is to evaluate the radiation dose reduction achieved using iterative reconstruction (IR) compared to filtered back projection (FBP) in coronary CT angiography (CCTA) and assess the impact on diagnostic image quality. A systematic search of seven electronic databases was performed to identify all studies using a developed keywords strategy. A total of 14 studies met the criteria and were included in a review analysis. The results showed that there was a significant reduction in radiation dose when using IR compared to FBP (P < 0.05). The mean and standard deviation (SD) difference of CTDIvol and dose-length-product (DLP) were 14.70 ± 6.87 mGy and 186 ± 120 mGy.cm respectively. The mean ± SD difference of effective dose (ED ) was 2.9 ± 1.7 mSv with the range from 1.0 to 5.0 mSv. The assessment of diagnostic image quality showed no significant difference (P > 0.05). The mean ± SD difference of image noise, signal-noise ratio (SNR) and contrast-noise ratio (CNR) were 1.05 ± 1.29 HU, 0.88 ± 0.56 and 0.63 ± 1.83 respectively. The mean ± SD percentages of overall image quality scores were 71.79 ± 12.29% (FBP) and 67.31 ± 22.96% (IR). The mean ± SD percentages of coronary segment analysis were 95.43 ± 2.57% (FBP) and 97.19 ± 2.62% (IR). In conclusion, this review analysis shows that CCTA with the use of IR leads to a significant reduction in radiation dose as compared to the use of FBP. Diagnostic image quality of IR at reduced dose (30-41%) is comparable to FBP at standard dose in the diagnosis of CAD.

Usefulness of Ultralow-Dose (Submillisievert) Chest CT Using Iterative Reconstruction for Initial Evaluation of Sharp Fish Bone Esophageal Foreign Body.AJR Am J Roentgenol. 2015;205:985-990. [PMID: 26496545 DOI: 10.2214/AJR.15.14353]

OBJECTIVE

The purpose of this article was to evaluate the usefulness of ultralow-dose chest CT as an initial imaging study for evaluation of sharp fish bone esophageal foreign body (FB).

MATERIALS AND METHODS

A total of 57 subjects who underwent ultralow-dose chest CT were included in this retrospective study. All subjects had a history of ingestion and symptoms of esophageal FB. All ultralow-dose chest CT data were reconstructed twice, once with filtered back projection (FBP) and once with iterative reconstruction, and three observers reviewed the images independently. ROC analysis was used to evaluate diagnostic performance of ultralow-dose chest CT. Intraclass correlation coefficient (ICC) was calculated for analysis of interobserver agreement.

RESULTS

Among 57 patients, 42 were confirmed as having esophageal FB. Significant objective noise reduction of mediastinum was achieved using an iterative reconstruction technique. Subjective image noise of iterative reconstruction was significantly better than that of FBP. Overall diagnostic performance of ultralow-dose chest CT for esophageal FB of iterative reconstruction (AUC = 0.999) was significantly better than that of FBP (AUC = 0.95) (p = 0.02). Interobserver agreement was greater for iterative reconstruction (ICC = 0.944) than for FBP (ICC = 0.778).

CONCLUSION

Ultralow-dose chest CT using iterative reconstruction provided satisfactory diagnostic image quality for identifying fish bone esophageal FB with reduced radiation dose and high observer accuracy. Therefore, ultralow-dose chest CT would be adequate as a first-line imaging modality for fish bone esophageal FB.

Effect of reconstruction algorithms on the accuracy of (99m)Tc sestamibi SPECT/CT parathyroid imaging

The superiority of SPECT/CT over SPECT for (99m)Tc-sestamibi parathyroid imaging often is assumed to be due to improved lesion localization provided by the anatomic component (computed tomography) of the examination. It also is possible that this superiority may be related to the algorithms used for SPECT data reconstruction. The objective of this investigation was to determine the effect of SPECT reconstruction algorithms on the accuracy of MIBI SPECT/CT parathyroid imaging. We retrospectively analyzed preoperative MIBI SPECT/CT parathyroid imaging studies performed on 106 patients. SPECT data were reconstructed by filtered back projection (FBP) and by iterative reconstruction with corrections for collimator resolution recovery and attenuation (IRC). Two experienced readers independently graded lesion detection certainty on a 5-point scale without knowledge of each other's readings, reconstruction methods, other test results or final diagnoses. All patients had surgical confirmation of the final diagnosis, including disease limited to the neck, and location and weight of excised lesion(s). There were 135 parathyroid lesions among the 106 patients. For FBP SPECT/CT and IRC SPECT/CT sensitivity was 76% and 90% (p = 0.003), specificity was 87% and 87% (p = 0.90), and accuracy was 83% and 88% (p = 0.04), respectively. Inter-rater agreement was significantly higher for IRC than for FBP (kappa = 0.76, "good agreement", versus kappa = 0.58, "moderate agreement", p < 0.0001). We conclude that the improved accuracy of MIBI SPECT/CT compared to MIBI SPECT for preoperative parathyroid lesion localization is due in part to the use of IRC for SPECT data reconstruction.

Quantitative analysis of the effect of iterative reconstruction using a phantom: determining the appropriate blending percentage

PURPOSE

To investigate the optimal blending percentage of adaptive statistical iterative reconstruction (ASIR) in a reduced radiation dose while preserving a degree of image quality and texture that is similar to that of standard-dose computed tomography (CT).

MATERIALS AND METHODS

The CT performance phantom was scanned with standard and dose reduction protocols including reduced mAs or kVp. Image quality parameters including noise, spatial, and low-contrast resolution, as well as image texture, were quantitatively evaluated after applying various blending percentages of ASIR. The optimal blending percentage of ASIR that preserved image quality and texture compared to standard dose CT was investigated in each radiation dose reduction protocol.

RESULTS

As the percentage of ASIR increased, noise and spatial-resolution decreased, whereas low-contrast resolution increased. In the texture analysis, an increasing percentage of ASIR resulted in an increase of angular second moment, inverse difference moment, and correlation and in a decrease of contrast and entropy. The 20% and 40% dose reduction protocols with 20% and 40% ASIR blending, respectively, resulted in an optimal quality of images with preservation of the image texture.

CONCLUSION

Blending the 40% ASIR to the 40% reduced tube-current product can maximize radiation dose reduction and preserve adequate image quality and texture.

Submillisievert chest CT with filtered back projection and iterative reconstruction techniques

OBJECTIVE

The purpose of this study was to compare submillisievert chest CT images reconstructed with filtered back projection (FBP), SafeCT, adaptive statistical iterative reconstruction (ASIR), and model-based iterative reconstruction (MBIR) with standard of care FBP images.

SUBJECTS AND METHODS

Fifty patients (33 men and 17 women; mean age [± SD], 62 ± 10 years) undergoing routine chest CT gave written informed consent for acquisition of an additional submillisievert chest CT series with reduced tube current but identical scanning length as standard of care chest CT. Sinogram data of the submillisievert series were reconstructed with FBP, SafeCT, ASIR, and MBIR and compared with FBP images at standard-dose chest CT (n = 8 × 50 = 400 series). Two thoracic radiologists performed independent comparison for visualization of lesion margin, visibility of small structures, and diagnostic acceptability. Objective noise measurements and noise spectral density were obtained.

RESULTS

Of 287 detected lesions, 162 were less than 1-cm noncalcified nodules. Lesion margins were well seen on all submillisievert reconstruction images except MBIR, on which they were poorly visualized. Likewise, only submillisievert MBIR images were suboptimal for visibility of normal structures, such as pulmonary vessels in the outer 2 cm of the lung, interlobular fissures, and subsegmental bronchial walls. MBIR had the lowest image noise compared with other techniques.

CONCLUSION

FBP, SafeCT, ASIR, and MBIR can enable optimal lesion evaluation on chest CT acquired at a volume CT dose index of 2 mGy. However, all submillisievert reconstruction techniques were suboptimal for visualization of mediastinal structures. Submillisievert MBIR images were suboptimal for visibility of normal lung structures despite showing lower image noise.

Ultra-low-dose CT of the lung: effect of iterative reconstruction techniques on image quality

RATIONALE AND OBJECTIVES

To compare quality of ultra-low-dose thin-section computed tomography (CT) images of the lung reconstructed using model-based iterative reconstruction (MBIR) and adaptive statistical iterative reconstruction (ASIR) to filtered back projection (FBP) and to determine the minimum tube current-time product on MBIR images by comparing to standard-dose FBP images.

MATERIALS AND METHODS

Ten cadaveric lungs were scanned using 120 kVp and four different tube current-time products (8, 16, 32, and 80 mAs). Thin-section images were reconstructed using MBIR, three ASIR blends (30%, 60%, and 90%), and FBP. Using the 8-mAs data, side-to-side comparison of the four iterative reconstruction image sets to FBP was performed by two independent observers who evaluated normal and abnormal findings, subjective image noise, streak artifact, and overall image quality. Image noise was also measured quantitatively. Subsequently, 8-, 16-, and 32-mAs MBIR images were compared to standard-dose FBP images. Comparisons of image sets were analyzed using the Wilcoxon signed rank test with Bonferroni correction.

RESULTS

At 8 mAs, MBIR images were significantly better (P < .005) than other reconstruction techniques except in evaluation of interlobular septal thickening. Each set of low-dose MBIR images had significantly lower (P < .001) subjective and objective noise and streak artifacts than standard-dose FBP images. Conspicuity and visibility of normal and abnormal findings were not significantly different between 16-mAs MBIR and 80-mAs FBP images except in identification of intralobular reticular opacities.

CONCLUSIONS

MBIR imaging shows higher overall quality with lower noise and streak artifacts than ASIR or FBP imaging, resulting in nearly 80% dose reduction without any degradations of overall image quality.