Diagnostic Performance of Dual-Layer Computed Tomography for Deep Vein Thrombosis in Indirect Computed Tomography Venography.

BACKGROUND: The aim of this study was to evaluate the quality and diagnostic performance of virtual monochromatic images (VMI) obtained with dual-layer dual-energy computed tomography (DL-DECT) during indirect CT venography (CTV) for deep vein thrombosis (DVT).Methods and Results:This retrospective study was approved by the Institutional Review Board, which waived the requirement for informed consent. We retrospectively enrolled 45 patients who underwent CTV with DL-DECT, and VMI were retrospectively generated. We compared the venous attenuation, noise, contrast, and contrast-to-noise ratio (CNR) between VMI with the highest CNR and conventional CT on paired t-test. Furthermore, we compared the pooled area under the curve (AUC) of each technique with Delong's test in 34 patients who underwent color Doppler ultrasonography. The 40-keV VMI had the best CNR. The noise was significantly lower on 40-keV (9.7±2.5 HU) than on 120-kVp VMI (10.5±2.5 HU; P<0.01). The contrast (120 kVp, 38.2±15.3 HU vs. 40 keV, 131.6±43.6 HU) and CNR (120 kVp, 3.8±1.7 vs. 40 keV, 14.4±6.1) were significantly higher in 40-keV VMI than in 120-kVp VMI (P<0.01). Furthermore, the pooled AUC was significantly higher for 40-keV (0.84) than for 120-kVp VMI (0.78; P=0.03). CONCLUSIONS: In indirect CTV, 40-keV VMI obtained with DL-DECT offers better image quality and diagnostic performance for DVT than conventional CT.

Volumetric modulated arc therapy treatment planning based on virtual monochromatic images for head and neck cancer: effect of the contrast-enhanced agent on dose distribution.

Virtual monochromatic images (VMIs) at a lower energy level can improve image quality but the computed tomography (CT) number of iodine contained in the contrast-enhanced agent is dramatically increased. We assessed the effect of the use of contrast-enhanced agent on the dose distributions in volumetric modulated arc therapy (VMAT) planning for head and neck cancer (HNC). Based on the VMIs at 40 keV (VMI40keV ), 60 keV(VMI60keV ), and 77 keV (VMI77keV ) of a tissue characterization phantom, lookup tables (LUTs) were created. VMAT plans were generated for 15 HNC patients based on contrast-enhanced- (CE-) VMIs at 40-, 60-, and 77 keV using the corresponding LUTs, and the doses were recalculated based on the noncontrast-enhanced- (nCE-) VMIs. For all structures, the difference in CT numbers owing to the contrast-enhanced agent was prominent as the energy level of the VMI decreased, and the mean differences in CT number between CE- and nCE-VMI was the largest for the clinical target volume (CTV) (125.3, 55.9, and 33.1 HU for VMI40keV , VMI60keV , and VMI77keV, respectively). The mean difference of the dosimetric parameters (D99% , D50% , D1% , Dmean , and D0.1cc ) for CTV and OARs was <1% in the treatment plans based on all VMIs. The maximum difference was observed for CTV in VMI40keV (2.4%), VMI60keV (1.9%), and VMI77keV (1.5%) plans. The effect of the contrast-enhanced agent was larger in the VMAT plans based on the VMI at a lower energy level for HNC patients. This effect is not desirable in a treatment planning procedure.

Reduction of metal artifacts from unilateral hip arthroplasty on dual-energy CT with metal artifact reduction software.

Background The evaluation of hip arthroplasty is a challenge in computed tomography (CT). The virtual monochromatic spectral (VMS) images with metal artifact reduction software (MARs) in spectral CT can reduce the artifacts and improve the image quality. Purpose To evaluate the effects of VMS images and MARs for metal artifact reduction in patients with unilateral hip arthroplasty. Material and Methods Thirty-five patients underwent dual-energy CT. Four sets of VMS images without MARs and four sets of VMS images with MARs were obtained. Artifact index (AI), CT number, and SD value were assessed at the periprosthetic region and the pelvic organs. The scores of two observers for different images and the inter-observer agreement were evaluated. Results The AIs in 120 and 140 keV images were significantly lower than those in 80 and 100 keV images. The AIs of the periprosthetic region in VMS images with MARs were significantly lower than those in VMS images without MARs, while the AIs of pelvic organs were not significantly different. VMS images with MARs improved the accuracy of CT numbers for the periprosthetic region. The inter-observer agreements were good for all the images. VMS images with MARs at 120 and 140 keV had higher subjective scores and could improve the image quality, leading to reliable diagnosis of prosthesis-related problems. Conclusion VMS images with MARs at 120 and 140 keV could significantly reduce the artifacts from hip arthroplasty and improve the image quality at the periprosthetic region but had no obvious advantage for pelvic organs.

Advanced virtual monochromatic reconstruction of dual-energy unenhanced brain computed tomography in children: comparison of image quality against standard mono-energetic images and conventional polychromatic computed tomography.

BACKGROUND: Advanced virtual monochromatic reconstruction from dual-energy brain CT has not been evaluated in children. OBJECTIVE: To determine the most effective advanced virtual monochromatic imaging energy level for maximizing pediatric brain parenchymal image quality in dual-energy unenhanced brain CT and to compare this technique with conventional monochromatic reconstruction and polychromatic scanning. MATERIALS AND METHODS: Using both conventional (Mono) and advanced monochromatic reconstruction (Mono+) techniques, we retrospectively reconstructed 13 virtual monochromatic imaging energy levels from 40 keV to 100 keV in 5-keV increments from dual-source, dual-energy unenhanced brain CT scans obtained in 23 children. We analyzed gray and white matter noise ratios, signal-to-noise ratios and contrast-to-noise ratio, and posterior fossa artifact. We chose the optimal mono-energetic levels and compared them with conventional CT. RESULTS: For Mono+maximum optima were observed at 60 keV, and minimum posterior fossa artifact at 70 keV. For Mono, optima were at 65-70 keV, with minimum posterior fossa artifact at 75 keV. Mono+ was superior to Mono and to polychromatic CT for image-quality measures. Subjective analysis rated Mono+superior to other image sets. CONCLUSION: Optimal virtual monochromatic imaging using Mono+ algorithm demonstrated better image quality for gray-white matter differentiation and reduction of the artifact in the posterior fossa.

Metal artifact reduction using virtual monochromatic images for patients with pedicle screws implants on CT.

PURPOSE: To evaluate the effect of spectral CT for metal artifact reduction in patients with pedicle screw. METHODS: 45 patients with 119 pairs of pedicle screws underwent spectral CT examination. One set of conventional (140 kVp) polychromatic image and nine sets of virtual monochromatic images (60-140 keV) were obtained. The standard deviation (SD) of CT number in 12 locations around the implant and 1 on homogenous fat was measured to generate artifact index (AI). Objective assessment including AI, CT number and SD value was performed with independent t test and paired sample t test. Two radiologists independently reviewed the image quality, and Wilcoxon signed-rank test and kappa analysis were used for the subjective scores of image quality. RESULTS: The artifact index (AI) of all the regions decreased as keV increased. AIs of 100-140 keV were lower than that of 140 kVp images. At 120 keV there was no significant difference in CT numbers of psoas major muscle and vertebral canal between pedicle screw level and pedicle level, but a significant difference in SD value was determined between the two levels. The subjective scores at 100-140 keV were higher than the images at 140 kVp, and the highest subjective score of two observers and excellent interobserver agreement were found at 120 keV (κ = 0.889). CONCLUSIONS: Virtual monochromatic images at high-energy levels have a well-concordant effect of removing metal artifacts, and 120 keV monochromatic images provided an accurate CT number and good subjective score.

Clinical Feasibility of Dual-Layer CT With Virtual Monochromatic Image for Preoperative Staging in Patients With Breast Cancer: A Comparison With Breast MRI.

OBJECTIVE: Dual-layer CT (DLCT) can create virtual monochromatic images (VMIs) at various monochromatic X-ray energies, particularly at low keV levels, with high contrast-to-noise ratio. The purpose of this study was to assess the clinical feasibility of contrast-enhanced chest DLCT with a low keV VMI for preoperative breast cancer staging, in comparison to breast MRI. MATERIALS AND METHODS: A total of 152 patients with 155 index breast cancers were enrolled in the study. VMIs were generated from contrast-enhanced chest DLCT at 40 keV and maximum intensity projection (MIP) with three-dimensional (3D) reconstruction was performed for both bilateral breast areas. Two radiologists reviewed in consensus the 3D MIP images of the chest DLCT with VMI and breast MRI in separate sessions with a 3-month wash-out period. The detection rate and mean tumor size of the index cancer were compared between the chest DLCT with VMI and breast MRI. Additionally, the agreement of tumor size measurement between the two imaging modalities were evaluated. RESULTS: Of all index cancers, 84.5% (131/155) were detected in the chest DLCT with VMI, while 88.4% (137/155) were detected in the breast MRI (P = 0.210). The Bland-Altman agreement between the chest DLCT with VMI and breast MRI was a mean difference of -0.05 cm with 95% limits of agreement of -1.29 to 1.19 cm. The tumor size in the chest DLCT with VMI (2.3 ± 1.7 cm) was not significantly different from that in the breast MRI (2.4 ± 1.6 cm) (P = 0.106). CONCLUSION: The feasibility of chest DLCT with VMI was demonstrated for preoperative tumor staging in breast cancer patients, showing comparable cancer detectability and good agreement in tumor size measurement compared to breast MRI. This suggests that chest DLCT with VMI can serve as a potential alternative for patients who have contraindications to breast MRI.

Optimal virtual monochromatic images for assessing metastatic lateral cervical lymph nodes in patients with papillary thyroid carcinoma using dual­layer spectral detector computed tomography.

PURPOSE: To determine the optimal virtual monochromatic images (VMIs) from dual-layer spectral detector computed tomography for the visualization and diagnosis of metastatic lateral cervical lymph nodes (LNs) in patients with papillary thyroid carcinoma (PTC). METHODS: Ninety-five lateral cervical LNs (49 metastatic and 46 non-metastatic) derived from 24 patients (16 females; mean age, 40.0 ± 13.4 years) were included. 40-100 kiloelectron voltage (keV) VMIs, 120 keV VMI and conventional 120 kV peak (kVp) polyenergetic image (PI) were reconstructed. Five-point scale of subjective image quality, signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) of LNs were assessed and compared among each VMI and 120 kVp PI. Receiver operating characteristic (ROC) curves and Delong tests were used to assess and compare the diagnostic efficacy of arterial enhancement fraction (AEF) based on each VMI and 120 kVp PI. RESULTS: 40 keV VMI showed significantly higher SNR and CNR in both arterial and venous phases, and better image quality in arterial phase than 70-100 keV VMIs, 120 keV VMI, and 120 kVp PI (all p < 0.05). In all sets of images, AEF values of metastatic LNs were significantly higher than those of non-metastatic LNs (all p < 0.05). When using AEF value of 40 keV VMI to diagnose metastatic lateral cervical LNs, an area under ROC curve (AUC) of 0.878, sensitivity of 87.8 % and specificity of 80.4 % could be obtained, while the AUC of AEF value of 120 kVp PI was 0.815 (p = 0.154). CONCLUSION: 40 keV VMI might be optimal for displaying and diagnosing the metastatic lateral cervical LNs in patients with PTC.

Added value of 40 keV virtual monoenergetic images for diagnosing malignant pleural effusion on chest CT.

OBJECTIVE: This study aimed to evaluate the added value of 40 keV virtual mono-energetic images (VMIs) obtained from dual-layer detector CT (DLCT) for diagnosing malignant pleural effusion (MPE) in patients presenting with unilateral pleural effusion on chest CT. MATERIALS AND METHODS: This retrospective study included 75 patients with unilateral pleural effusion who underwent contrast-enhanced chest CT scans using DLCT. Quantitative and qualitative assessments of the visibility of pleural thickening were conducted on both conventional 120 kVp images and 40 keV VMIs. Two independent radiologists reviewed chest CT scans with or without 40 keV VMIs to detect pleural nodules or nodular thickening for the diagnosis of MPE. Diagnostic performances were compared and independent predictors of MPE were identified through multivariate logistic regression analysis using CT and clinicopathologic findings. RESULTS: Pleural thickening associated with MPE demonstrated a higher contrast-to-noise ratio value and greater visual conspicuity in 40 keV VMIs compared to benign effusions (p < 0.05). For both readers, the use of 40 keV VMIs significantly improved (p < 0.05) the diagnostic performance in terms of sensitivity and area under the curve (AUC) for diagnosing MPE through the detection of pleural nodularity. Inter-observer agreements between the two readers were substantial for both 120 kVp images alone and the combined use of 40 keV VMIs. Initial cytology results and pleural nodularity at 40 keV were identified as independent predictors of MPE. CONCLUSION: The use of 40 keV VMIs from DLCT can improve diagnostic performance of readers in detecting MPE among patients with unilateral pleural effusion.

Utility of spectral CT with orthopedic metal artifact reduction algorithms for 125I seeds implantation in mediastinal and hepatic tumors.

Background: Virtual monochromatic image (VMI) combined with orthopedic metal artifact reduction algorithms (VMI + O-MAR) can effectively reduce artifacts caused by metal implants of different types. Nevertheless, so far, no study has systematically evaluated the efficacy of VMI + O-MAR in reducing various types of metal artifacts induced by 125I seeds. The aim of this study was to assess the effectiveness of combining spectral computed tomography (CT) images with O-MAR in reducing metal artifacts and improving the image quality affected by artifacts in patients after 125I radioactive seeds implantation (RSI). Methods: A total of 45 patients who underwent dual-layer detector spectral CT (DLCT; IQon, Philips Healthcare) scanning of mediastinal and hepatic tumors after 125I RSI were retrospectively included. Spectral data were reconstructed into conventional image (CI), VMI, CI combined with O-MAR (CI + O-MAR), and VMI + O-MAR to evaluate the de-artifact effect and image quality improvement. Objective indicators included signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), and artifact index (AI) of lesions affected by artifacts. Subjective indicators included assessment of overcorrected artifacts and new artifacts, different morphology of artifacts, and overall image quality. Results: In artifact-affected lesion areas, SNR and CNR in the CI/VMI + O-MAR groups were better than those in CI groups (all P values <0.05). The AI showed a downward trend as VMI keV increased (all P values <0.001). The AI values of the CI/VMI (50-150 keV) group were all higher than the groups of CI/VMI + O-MAR (50-150 keV) (P<0.001). Overcorrection artifacts and new artifacts were concentrated in the VMI50/70 keV groups. In the evaluation of artifact morphology, as the VMI keV increased, the number of near-field banding artifacts in hyperdense artifacts gradually decreased, whereas the number of minimal or no artifacts increased, and the total number of hyperdense artifacts were decreased. The diagnostic and image quality scores of hyperdense artifacts were higher than those of hypodense artifacts as VMI keV increased. Conclusions: High VMI level combined with O-MAR substantially improve objective and subjective image quality, lesion display ability, and diagnostic confidence of CT follow-up after 125I RSI, especially at the VMI + O-MAR 150 keV level.

Virtual monochromatic spectral CT imaging in preoperative evaluations for intraductal spread of breast cancer: comparison with conventional CT and MRI.

PURPOSE: To investigate the efficacy of virtual monochromatic spectral computed tomography imaging (VMI) in the preoperative evaluation for intraductal spread of breast cancer. MATERIALS AND METHODS: Twenty-four women who underwent spectral CT and were pathologically diagnosed with ductal carcinoma with a ≥ 2-cm noninvasive component were retrospectively enrolled in Group 1. Twenty-two women with 22 lesions pathologically diagnosed with ductal carcinoma in situ or microinvasive carcinoma were enrolled in Group 2. We compared the contrast-to-noise ratios (CNRs) of the lesions on conventional 120-kVp CT images and 40-keV VMIs in Group 1. Two board-certified radiologists measured the maximum diameters of enhancing areas on 120-kVp CT, 40-keV VMI, and MRI in Group 2 and compared with histopathological sizes. RESULTS: The quantitative assessment of Group 1 revealed that the mean ± SD of the CNRs in the 40-keV images were significantly greater than those in the 120-kVp images (5.5 ± 1.9 vs. 3.6 ± 1.5, p < 0.0001). The quantitative assessment of Group 2 demonstrated that the lesion size observed in the conventional 120-kVp CT images by both readers was significantly underestimated as compared to the histopathological size (p = 0.017, 0.048), whereas both readers identified no significant differences between the lesion size measured on 40-keV VMI and the histopathological data. In a comparison with MRI, 40-keV VMI provided measurement within a 10-mm error range in more lesions as compared to the conventional 120-kVp CT. CONCLUSION: VMI improves the evaluation of intraductal spread and is useful for the preoperative evaluations of breast cancer.

Pseudo low-energy monochromatic imaging of head and neck cancers: Deep learning image reconstruction with dual-energy CT.

PURPOSE: Low-energy virtual monochromatic images (VMIs) derived from dual-energy computed tomography (DECT) systems improve lesion conspicuity of head and neck cancer over single-energy CT (SECT). However, DECT systems are installed in a limited number of facilities; thus, only a few facilities benefit from VMIs. In this work, we present a deep learning (DL) architecture suitable for generating pseudo low-energy VMIs of head and neck cancers for facilities that employ SECT imaging. METHODS: We retrospectively analyzed 115 patients with head and neck cancers who underwent contrast enhanced DECT. VMIs at 70 and 50 keV were used as the input and ground truth (GT), respectively. We divided them into two datasets: for DL (104 patients) and for inference with SECT (11 patients). We compared four DL architectures: U-Net, DenseNet-based, and two ResNet-based models. Pseudo VMIs at 50 keV (pVMI50keV) were compared with the GT in terms of the mean absolute error (MAE) of Hounsfield unit (HU) values, peak signal-to-noise ratio (PSNR), and structural similarity (SSIM). The HU values for tumors, vessels, parotid glands, muscle, fat, and bone were evaluated. pVMI50keV were generated from actual SECT images and the HU values were evaluated. RESULTS: U-Net produced the lowest MAE (13.32 ± 2.20 HU) and highest PSNR (47.03 ± 2.33 dB) and SSIM (0.9965 ± 0.0009), with statistically significant differences (P < 0.001). The HU evaluation showed good agreement between the GT and U-Net. U-Net produced the smallest absolute HU difference for the tumor, at < 5.0 HU. CONCLUSION: Quantitative comparisons of physical parameters demonstrated that the proposed U-Net could generate high accuracy pVMI50keV in a shorter time compared with the established DL architectures. Although further evaluation on diagnostic accuracy is required, our method can help obtain low-energy VMI from SECT images without DECT systems.

Accuracy of virtual monochromatic images generated by the decomposition of photoelectric absorption and Compton scatter in dual-energy computed tomography.

The decomposition of the linear attenuation coefficient into photoelectric absorption and Compton scattering provides virtual monochromatic images (VMIs). The accuracy of the computed tomography (CT) number of VMI, which is obtained by decomposing the linear attenuation coefficient into photoelectric absorption and Compton scattering, was verified in the energy range of 40-200 keV. The possibility of improving the accuracy of CT numbers by using pre-energy-calibrated images as input was also investigated. The VMIs were generated in two groups of images: (i) dual-energy scanned images and (ii) high- and low-energy images generated by two-material decomposition (i.e., pre-energy-calibrated images). The object for analysis was solid iodine rods inserted in the center of the multi-energy CT phantom. The VMIs were generated from the dual-energy scanned images and pre-energy-calibrated images, and the theoretical and measured CT numbers of solid iodine rods were compared. Furthermore, the absolute error (AE) and relative error (RE) were calculated. With both images, the accuracy of the CT numbers was extremely high for regions close to the high- and low-tube-voltage X-ray energy or the high and low energy of the input images. By using the pre-energy-calibrated images, the maximum AE was reduced from 133 to 96 HU at an energy of 40 keV. Similarly, the maximum RE was reduced from 325 to 50% at an energy of 200 keV. The pre-energy-calibrated images reduced the overall error of the CT numbers and controlled the energy region where accurate CT numbers could be obtained.

Dual-layer spectral detector CT for contrast agent concentration, dose and injection rate reduction: Utility in imaging of the superior mesenteric artery.

PURPOSE: To investigate the feasibility of contrast agent and injection rate reduction for dual-layer spectral detector computed tomography (SDCT) imaging of the superior mesenteric artery (SMA) using virtual monochromatic image (VMI). METHODS: A total of 102 patients who underwent abdominal arterial phase-enhanced SDCT examination due to suspected abdominal diseases were prospectively selected and divided into control group, low concentration/dose groups (groups 370-0.7, 300-1.0, and 300-0.9) and low injection rate groups (groups 2-370 and 2-350). Compared with the control group, low concentration/dose groups and low injection rate groups lowered the concentration/dose or injection rate of the contrast agent to varying degrees. The raw data obtained in each group were reconstructed using hybrid-iterative reconstruction and projection spatial-spectral reconstruction algorithm. The image quality of the SMA in conventional images (CI) and in VMIs40-140 kiloelectron volt (keV) (interval: 10 keV) during the arterial phase was analyzed. Multiplanar reformation images and volume rendering images of the SMA were reconstructed. Image quality objective evaluation indexes included the CT values, contrast-to-noise ratio, signal-to-noise ratio, and diameter of the SMA. The diameter of the SMA was determined by the CT values profile curve and its full width at half maximum. Two doctors independently evaluated the subjective image quality of multiplanar reformation coronal images and volume rendering images according to a 5-point scale. Repeated analysis of variance and Friedman test were used to compare the differences in the objective evaluation indexes and subjective scores between VMIs and CI in the same group. The Dunnett's t-test or Dunnett's T3 test and Kruskal-Wallis H-test were used to compare the differences in the objective evaluation indexes and subjective scores between the experimental and control groups. RESULTS: VMIs of the SMA in each group had the best image quality at 60 keV, and VMI60 keV in each group were better than their respective CI to varying degrees. Although the objective (CT values, contrast-to-noise ratio, and signal-to-noise ratio) and subjective (subjective scores) indexes of CI in the low concentration/dose groups and low injection rate groups were lower than those of CI in the control group to varying degrees, these indexes of VMI60 keV in the low concentration/dose groups and group 2-370 were equal to or even better than the CI in the control group. CONCLUSIONS: VMI60 keV using SDCT could effectively reduce the contrast agent load while providing equivalent or better SMA image quality compared with CI obtained using a conventional contrast agent protocol. When the injection rate was lowered to 2.0 ml/s for a high-concentration contrast agent (370 mgI/ml), the SMA image quality at VMI60 keV was comparable with that of the CI in the control group.

Comparative study of the image quality of twin beam dual energy and single energy carotid CT angiography.

PURPOSE: The aim of this study was to compare the image quality of low virtual monochromatic image (VMI) from the twin-beam dual energy (TBDE) mode to that from the single energy (SE) mode using both standard and low tube voltage acquisition in carotid computed tomography angiography (CTA). METHODS: In this prospective study, 221 patients were imaged using TBDE mode with 120 kV or SE mode with 80/90/120 kV, patient numbers in all groups are equal(n = 55) except in the group using 80 kV SE mode(n = 56). VMIs ranging from 40 to 90 keV with a step of 1 keV were reconstructed from TBDE mode. In objective image quality assessment, regions of interest (ROIs) were placed in both carotid artery lumens and in both sternocleidomastoid muscles to calculate and compare the contrast-to-noise ratio (CNR) and the dose-normalized CNR (CNRD) among groups. In subjective assessment, a total of 13 arterial segments were assessed using a four-point Likert scale by two observers. Cohen's kappa test was used to quantify the level of agreement between the two observers. RESULTS: For ROI1 at the level of the carotid bifurcation, VMIs showed a higher CNR than the 120 kV SE group (p = 0.028)/ 90 kV SE group (p = 0.037) when their energy level were lower than 79 keV (56.92 ± 16.01)/56 keV (90.08 ± 22.14) respectively. The 90 kV SE (80.68 ± 24.47) showed the best CNR in all the SE groups. For CNRD, the 120/90/80 kV SE group was equivalent to 83/63/67 keV VMIs respectively. For ROI2 at the level of the origin of the common carotid artery, VMIs also showed a higher CNR than the 120 kV SE group (p = 0.015)/90 kV SE group (p = 0.034) when their energy level were lower than 83 keV (46.31 ± 14.47)/60 keV (72.23 ± 16.96) respectively. The 90 kV SE (64.98 ± 18.51) showed the best CNR in all the SE groups. For CNRD, the 120/90/80 kV SE group was equivalent to 88/72/75 keV VMIs respectively. The highest subjective rating score was 50 keV TBDE(3.70 ± 0.53). Cohen's Kappa values(0.79-0.85) suggest a substantial level of agreement between the two observers. CONCLUSIONS: The novel TBDE technique with low keV VMI reconstruction provides better image quality of carotid CTA than the low tube voltage scan in the SE mode. VMIs with a keV level below 56 keV have a higher CNR than those from SE scans with 80/90/120 kV. Subjectively, the optimal keV energy level in TBDE for carotid CT angiography is 50 keV.

Improvement of image quality for pancreatic cancer using deep learning-generated virtual monochromatic images: Comparison with single-energy computed tomography.

PURPOSE: To construct a deep convolutional neural network that generates virtual monochromatic images (VMIs) from single-energy computed tomography (SECT) images for improved pancreatic cancer imaging quality. MATERIALS AND METHODS: Fifty patients with pancreatic cancer underwent a dual-energy CT simulation and VMIs at 77 and 60 keV were reconstructed. A 2D deep densely connected convolutional neural network was modeled to learn the relationship between the VMIs at 77 (input) and 60 keV (ground-truth). Subsequently, VMIs were generated for 20 patients from SECT images using the trained deep learning model. RESULTS: The contrast-to-noise ratio was significantly improved (p < 0.001) in the generated VMIs (4.1 ± 1.8) compared to the SECT images (2.8 ± 1.1). The mean overall image quality (4.1 ± 0.6) and tumor enhancement (3.6 ± 0.6) in the generated VMIs assessed on a five-point scale were significantly higher (p < 0.001) than that in the SECT images (3.2 ± 0.4 and 2.8 ± 0.4 for overall image quality and tumor enhancement, respectively). CONCLUSIONS: The quality of the SECT image was significantly improved both objectively and subjectively using the proposed deep learning model for pancreatic tumors in radiotherapy.

Conditional generative adversarial networks to generate pseudo low monoenergetic CT image from a single-tube voltage CT scanner.

PURPOSE: To generate pseudo low monoenergetic CT images of the abdomen from 120-kVp CT images with cGAN. MATERIALS AND METHODS: We retrospectively included 48 patients who underwent contrast-enhanced abdominal CT using dual-energy CT. We reconstructed paired data sets of 120 kVp CT images and virtual low monoenergetic (55-keV) CT images. cGAN was prepared to generate pseudo 55-keV CT images from 120-kVp CT images. The pseudo 55 keV CT images in epoch 10, 50, 100, and 500 were compared to the 55 keV images generated using peak signal-to-noise ratio (PSNR) and structural similarity index (SSIM). RESULTS: The PSNRs were 28.0, 28.5, 28.6, and 28.8 at epochs 10, 50, 100, and 500, respectively. The SSIM was approximately constant from epochs 50 to 500. CONCLUSION: Pseudo low monoenergetic abdominal CT images were generated from 120-kVp CT images using cGAN, and the images had good quality similar to that of monochromatic images obtained with DECT software.

A New Outlook on the Ability to Accumulate an Iodine Contrast Agent in Solid Lung Tumors Based on Virtual Monochromatic Images in Dual Energy Computed Tomography (DECT): Analysis in Two Phases of Contrast Enhancement.

For some time, dual energy computed tomography (DECT) has been an established method used in a vast array of clinical applications, including lung nodule assessment. The aim of this study was to analyze (using monochromatic DECT images) how the X-ray absorption of solitary pulmonary nodules (SPNs) depends on the iodine contrast agent and when X-ray absorption is no longer dependent on the accumulated contrast agent. Sixty-six patients with diagnosed solid lung tumors underwent DECT scans in the late arterial phase (AP) and venous phase (VP) between January 2017 and June 2018. Statistically significant correlations (p ≤ 0.001) of the iodine contrast concentration were found in the energy range of 40-90 keV in the AP phase and in the range of 40-80 keV in the VP phase. The strongest correlation was found between the concentrations of the contrast agent and the scanning energy of 40 keV. At the higher scanning energy, no significant correlations were found. We concluded that it is most useful to evaluate lung lesions in DECT virtual monochromatic images (VMIs) in the energy range of 40-80 keV. We recommend assessing SPNs in only one phase of contrast enhancement to reduce the absorbed radiation dose.

Volumetric modulated arc therapy planning based on virtual monochromatic images: Effect of inaccurate CT numbers on dose distributions.

PURPOSE: Though virtual monochromatic images (VMIs) at low energy levels can improve image quality, the measured Hounsfield unit (HU) values can be inaccurate. We assessed the dosimetric error due to inaccurate HU estimation in volumetric modulated arc therapy (VMAT) planning. METHODS: Based on the VMIs at 50 keV (VMI50keV), 77 keV (VMI77keV) and single-energy CT (SECT) image for a phantom with different sizes, lookup tables (LUTL and LUTS) were created. Using an anthropomorphic phantom (head and spine regions), VMAT plans were generated based on VMI50keV, VMI77keV and SECT using the corresponding LUTL, and then, the doses were re-calculated using LUTS. For clinical cases, 30 VMAT plans (prostate, brain, and spine cases) were generated based on VMI50keV and VMI77keV. RESULTS: In the anthropomorphic phantom study, the difference in the dosimetric parameters for planning target volume (PTV) in the VMAT plan based on the VMI77keV was smallest (within 0.1 Gy) among three types of treatment planning approach. In clinical cases, in general, the differences of the 3-dimensional gamma passing rate and dosimetric parameters in the treatment plans based on the VMI50keV were larger than those in the VMI77keV. Especially for brain cases, the difference for PTV was more prominent when AXB was used (the maximum difference was 0.5 Gy) than AAA. CONCLUSIONS: The dosimetric error due to the inaccurate HU estimation was larger in the VMIs at low energy levels. This may be clinically insignificant, but should be avoided in the VMAT treatment planning.

Dual-Energy CT Texture Analysis With Machine Learning for the Evaluation and Characterization of Cervical Lymphadenopathy.

PURPOSE: To determine whether machine learning assisted-texture analysis of multi-energy virtual monochromatic image (VMI) datasets from dual-energy CT (DECT) can be used to differentiate metastatic head and neck squamous cell carcinoma (HNSCC) lymph nodes from lymphoma, inflammatory, or normal lymph nodes. MATERIALS AND METHODS: A retrospective evaluation of 412 cervical nodes from 5 different patient groups (50 patients in total) having undergone DECT of the neck between 2013 and 2015 was performed: (1) HNSCC with pathology proven metastatic adenopathy, (2) HNSCC with pathology proven benign nodes (controls for (1)), (3) lymphoma, (4) inflammatory, and (5) normal nodes (controls for (3) and (4)). Texture analysis was performed with TexRAD® software using two independent sets of contours to assess the impact of inter-rater variation. Two machine learning algorithms (Random Forests (RF) and Gradient Boosting Machine (GBM)) were used with independent training and testing sets and determination of accuracy, sensitivity, specificity, PPV, NPV, and AUC. RESULTS: In the independent testing (prediction) sets, the accuracy for distinguishing different groups of pathologic nodes or normal nodes ranged between 80 and 95%. The models generated using texture data extracted from the independent contour sets had substantial to almost perfect agreement. The accuracy, sensitivity, specificity, PPV, and NPV for correctly classifying a lymph node as malignant (i.e. metastatic HNSCC or lymphoma) versus benign were 92%, 91%, 93%, 95%, 87%, respectively. CONCLUSION: Machine learning assisted-DECT texture analysis can help distinguish different nodal pathology and normal nodes with a high accuracy.