ABSTRACT
Background: In this study, we present a quantitative method to evaluate the motion artifact correction (MAC) technique through the morphological analysis of blood vessels in the images before and after MAC. Methods: Cone-beam computed tomography (CBCT) scans of 37 patients who underwent transcatheter chemoembolization were obtained, and images were reconstructed with and without the MAC technique. First, two interventional radiologists selected the blood vessels corrected by MAC. We devised a motion-corrected index (MCI) metric that analyzed the morphology of blood vessels in 3D space using information on the centerline of blood vessels, and the blood vessels selected by the interventional radiologists were quantitatively evaluated using MCI. In addition, these blood vessels were qualitatively evaluated by two interventional radiologists. To validate the effectiveness of the devised MCI, we compared the MCI values in a blood vessel corrected by MAC and one non-corrected by MAC. Results: The visual evaluation revealed that motion correction was found in the images of 23 of 37 patients (62.2%), and a performance evaluation of MAC was performed with 54 blood vessels in 23 patients. The visual grading analysis score was 1.56 ± 0.57 (radiologist 1) and 1.56 ± 0.63 (radiologist 2), and the proposed MCI was 0.67 ± 0.11, indicating that the vascular morphology was well corrected by the MAC. Conclusions: We verified that our proposed method is useful for evaluating the MAC technique of CBCT, and the MAC technique can correct the blood vessels distorted by the patient's movement and respiration.
ABSTRACT
For the comprehensive evaluation of metal artifact reduction (MAR) technique, not only the removal of metal artifacts but also the evaluation of the area restored by MAR is required. We propose a method to comprehensively evaluate the effect by MAR in this study. We have conducted the computed tomography scan to acquire both the evaluation image and the reference image for the full-reference based evaluation. The evaluation image and reference image were reconstructed into 24 image sets according to the tube potentials, image reconstruction method, and use of the MAR technique. Images of two different positions were selected according to the distance from metal and material (bone, tissue) distribution, and bone and tissue were automatically segmented in both evaluation and reference images. The values of full width at half the maximum (FWHM) and centroid were extracted after Gaussian modeling of each segmented region. Then, we computed four evaluation metrics (FWHMNM: non-MAR to non-metal ratio of FWHM, FWHMM: MAR to non-metal ratio of FWHM, CENTNM: non-MAR to non-metal ratio of centroid, CENTM: MAR to non-metal ratio of centroid), and the MAR image and non-MAR image were compared. The overlap ratio automatically segmented from the evaluation image and reference image were position 1 (bone: 99.61%, tissue: 99.23%) with 80 kVp, position 1 (bone: 99.32%, tissue: 99.56%) with 120 kVp, position 2 (bone: 99.20%, tissue: 99.73%) with 80 kVp, and position 2 (bone: 99.23%, tissue: 99.67%) with 120 kVp. The FWHMNM showing the change of image pixel value by metal artifact was calculated as (bone: 1.32-1.46, tissue: 1.08-1.16) at 80 kVp and (bone: 1.19-1.27, tissue: 1.02-1.05) at 120 kVp. More metal artifacts occurred at 80 kVp tube potential. Regardless of the tube potential and image reconstruction method, the MAR showed an overall artifact reduction effect (1 < FWHMM < FWHMNM). However, distortion of pixel values occurred due to the MAR in regions where metal artifacts were high in proximity to metal (1 < FWHMNM < FWHMM). Overall, the average value of the medium was maintained (CENTM: 0.98-1.03) after MAR application, but there was a change of image value in region around the metal (CENTM: 0.97-1.11). In this study, we propose a new method to evaluate the effect of metal artifacts and MAR technique using full-reference based method. Metal artifacts, effect of MAR technique, and side-effect caused by MAR technique were quantitatively analyzed through proposed method. There are some limitations in applying it to clinical imaging since our method is a reference-based evaluation. However, our experimental results were important for understanding the effects of the MAR technique and its functional properties.
Subject(s)
Artifacts , Image Processing, Computer-Assisted , Algorithms , Image Processing, Computer-Assisted/methods , Metals , Phantoms, Imaging , Tomography, X-Ray Computed/methodsABSTRACT
This study evaluated the feasibility of deep-learning-based image reconstruction (DLIR) on coronary computed tomography angiography (CCTA). By using a 20 cm water phantom, the noise reduction ratio and noise power spectrum were evaluated according to the different reconstruction methods. Then 46 patients who underwent CCTA were retrospectively enrolled. CCTA was performed using the 16 cm coverage axial volume scan technique. All CT images were reconstructed using filtered back projection (FBP); three model-based iterative reconstructions (MBIR) of 40%, 60%, and 80%; and three DLIR algorithms: low (L), medium (M), and high (H). Quantitative and qualitative image qualities of CCTA were compared according to the reconstruction methods. In the phantom study, the noise reduction ratios of MBIR-40%, MBIR-60%, MBIR-80%, DLIR-L, DLIR-M, and DLIR-H were 26.7 ± 0.2%, 39.5 ± 0.5%, 51.7 ± 0.4%, 33.1 ± 0.8%, 43.2 ± 0.8%, and 53.5 ± 0.1%, respectively. The pattern of the noise power spectrum of the DLIR images was more similar to FBP images than MBIR images. In a CCTA study, CCTA yielded a significantly lower noise index with DLIR-H reconstruction than with the other reconstruction methods. DLIR-H showed a higher SNR and CNR than MBIR (p < 0.05). The qualitative image quality of CCTA with DLIR-H was significantly higher than that of MBIR-80% or FBP. The DLIR algorithm was feasible and yielded a better image quality than the FBP or MBIR algorithms on CCTA.
ABSTRACT
Background: The aim of this study was to compare the dose reduction potential and image quality of deep learning-based image reconstruction (DLIR) with those of filtered back-projection (FBP) and iterative reconstruction (IR) and to determine the clinically usable dose of DLIR for low-dose chest computed tomography (LDCT) scans. Methods: Multi-slice computed tomography (CT) scans of a chest phantom were performed with various tube voltages and tube currents, and the images were reconstructed using seven methods to control the amount of noise reduction: FBP, three stages of IR, and three stages of DLIR. For subjective image analysis, four radiologists compared 48 image data sets with reference images and rated on a 5-point scale. For quantitative image analysis, the signal to noise ratio (SNR), contrast to noise ratio (CNR), nodule volume, and nodule diameter were measured. Results: In the subjective analysis, DLIR-Low (0.46 mGy), DLIR-Medium (0.31 mGy), and DLIR-High (0.18 mGy) images showed similar quality to the FBP (2.47 mGy) image. Under the same dose conditions, the SNR and CNR were higher with DLIR-High than with FBP and all the IR methods (all P<0.05). The nodule volume and size with DLIR-High were significantly closer to the real volume than with FBP and all the IR methods (all P<0.001). Conclusions: DLIR can improve the image quality of LDCT compared to FBP and IR. In addition, the appropriate effective dose for LDCT would be 0.24 mGy with DLIR-High.
ABSTRACT
Purpose: To compare the image quality of CT obtained using a deep learning-based image reconstruction (DLIR) engine with images with adaptive statistical iterative reconstruction-V (AV). Materials and Methods: Using a phantom, the noise power spectrum (NPS) and task-based transfer function (TTF) were measured in images with different reconstructions (filtered back projection [FBP], AV30, 50, 100, DLIR-L, M, H) at multiple doses. One hundred and twenty abdominal CTs with 30% dose reduction were processed using AV30, AV50, DLIR-L, M, H. Objective and subjective analyses were performed. Results: The NPS peak of DLIR was lower than that of AV30 or AV50. Compared with AV30, the NPS average spatial frequencies were higher with DLIR-L or DLIR-M. For lower contrast objects, TTF in images with DLIR were higher than those with AV. The standard deviation in DLIR-H and DLIR-M was significantly lower than AV30 and AV50. The overall image quality was the best for DLIR-M (p < 0.001). Conclusions: DLIR showed improved image quality and decreased noise under a decreased radiation dose.
ABSTRACT
(1) Background: Highly flexible adaptive image receive (AIR) coil has become available for clinical use. The present study aimed to evaluate the performance of AIR anterior array coil in lung MR imaging using a zero echo time (ZTE) sequence compared with conventional anterior array (CAA) coil. (2) Methods: Sixty-six patients who underwent lung MR imaging using both AIR coil (ZTE-AIR) and CAA coil (ZTE-CAA) were enrolled. Image quality of ZTE-AIR and ZTE-CAA was quantified by calculating blur metric value, signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR) of lung parenchyma. Image quality was qualitatively assessed by two independent radiologists. Lesion detection capabilities for lung nodules and emphysema and/or lung cysts were evaluated. Patients' comfort levels during examinations were assessed. (3) Results: SNR and CNR of lung parenchyma were higher (both p < 0.001) in ZTE-AIR than in ZTE-CAA. Image sharpness was superior in ZTE-AIR (p < 0.001). Subjective image quality assessed by two independent readers was superior (all p < 0.05) in ZTE-AIR. AIR coil was preferred by 64 of 66 patients. ZTE-AIR showed higher (all p < 0.05) sensitivity for sub-centimeter nodules than ZTE-CAA by both readers. ZTE-AIR showed higher (all p < 0.05) sensitivity and accuracy for detecting emphysema and/or cysts than ZTE-CAA by both readers. (4) Conclusions: The use of highly flexible AIR coil in ZTE lung MR imaging can improve image quality and patient comfort. Application of AIR coil in parenchymal imaging has potential for improving delineation of low-density parenchymal lesions and tiny nodules.
ABSTRACT
OBJECTIVES: To compare the image quality and radiation dose of a deep learning image reconstruction (DLIR) algorithm compared with iterative reconstruction (IR) and filtered back projection (FBP) at different tube voltages and tube currents. MATERIALS AND METHODS: A customized body phantom was scanned at different tube voltages (120, 100, and 80 kVp) with different tube currents (200, 100, and 60 mA). The CT datasets were reconstructed with FBP, hybrid IR (30% and 50%), and DLIR (low, medium, and high levels). The reference image was set as an image taken with FBP at 120 kVp/200 mA. The image noise, contrast-to-noise ratio (CNR), sharpness, artifacts, and overall image quality were assessed in each scan both qualitatively and quantitatively. The radiation dose was also evaluated with the volume CT dose index (CTDIvol) for each dose scan. RESULTS: In qualitative and quantitative analyses, compared with reference images, low-dose CT with DLIR significantly reduced the noise and artifacts and improved the overall image quality, even with decreased sharpness (p < 0.05). Despite the reduction of image sharpness, low-dose CT with DLIR could maintain the image quality comparable to routine-dose CT with FBP, especially when using the medium strength level. CONCLUSION: The new DLIR algorithm reduced noise and artifacts and improved overall image quality, compared to FBP and hybrid IR. Despite reduced image sharpness in CT images of DLIR algorithms, low-dose CT with DLIR seems to have an overall greater potential for dose optimization. KEY POINTS: ⢠Using deep learning image reconstruction (DLIR) algorithms, image quality was maintained even with a radiation dose reduced by approximately 70%. ⢠DLIR algorithms yielded lower image noise, higher contrast-to-noise ratios, and higher overall image quality than FBP and hybrid IR, both subjectively and objectively. ⢠DLIR algorithms can provide a better image quality, much better than FBP and even better than hybrid IR, while facilitating a reduction in radiation dose.
Subject(s)
Deep Learning , Radiographic Image Interpretation, Computer-Assisted , Algorithms , Humans , Image Processing, Computer-Assisted , Multidetector Computed Tomography , Radiation Dosage , Radiographic Image Interpretation, Computer-Assisted/methodsABSTRACT
OBJECTIVES: To compare image noise and sharpness of vessels, liver, and muscle in lower extremity CT angiography between "adaptive statistical iterative reconstruction-V" (ASIR-V) and deep learning reconstruction "TrueFidelity" (TFI). METHODS: Thirty-seven patients (mean age, 65.2 years; 32 men) with lower extremity CT angiography were enrolled between November and December 2019. Images were reconstructed with two ASIR-V (blending factor of 80% and 100% (AV-100)) and three TFI (low-, medium-, and high-strength-level (TF-H) settings). Two radiologists evaluated these images for vessels (aorta, femoral artery, and popliteal artery), liver, and psoas muscle. For quantitative analyses, conventional indicators (CT number, image noise, signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR)) and blur metric values (indicating the degree of image sharpness) of selected regions of interest were determined. For qualitative analyses, the degrees of quantum mottle and blurring were assessed. RESULTS: The higher the blending factor in ASIR-V or the strength in TFI, the lower the noise, the higher the SNR and CNR values, and the higher the blur metric values in all structures. The SNR and CNR values of TF-H images were significantly higher than those of AV-80 images and similar to those of AV-100 images. The blur metric values in TFI images were significantly lower than those in ASIR-V images (p < 0.001), indicating increased sharpness. Among all the investigated image procedures, the overall qualitative image quality was best in TF-H images. CONCLUSION: TF-H was the most balanced image in terms of image noise and sharpness among the examined image combinations. KEY POINTS: ⢠Deep learning image reconstruction "TrueFidelity" is superior to iterative reconstruction "ASIR-V" regarding image noise and sharpness. ⢠The high-strength "TrueFidelity" approach generated the best image quality among the examined image reconstruction procedures. ⢠In iterative and deep learning CT image reconstruction, the higher the blending and strength factors, the lower the image noise and the poorer the image sharpness.
Subject(s)
Deep Learning , Aged , Algorithms , Humans , Male , Radiation Dosage , Radiographic Image Interpretation, Computer-Assisted , Signal-To-Noise Ratio , Tomography, X-Ray ComputedABSTRACT
PURPOSE: To evaluate the effects of gemstone spectral imaging-metal artifact reduction (GSI-MAR) on different dual-energy CT monochromatic images for patients with total knee replacement arthroplasty (TKRA) and to identify an appropriate protocol for clinical practice. METHOD: We enrolled 34 patients with TKRA. CT images were iteratively reconstructed with or without GSI-MAR at different energy levels (70, 95, 115, and 140â¯keV). Two radiologists evaluated the objective and subjective image qualities and MAR-related new artifacts at the femoral and tibial levels. For objective analysis, the mean CT number and image noise of the selected regions of interest in the bone and muscle were recorded. To quantitatively evaluate the performance of GSI-MAR, a structural similarity index (SSIM) was used. For subjective analysis, streak artifacts and diagnostic confidence in detecting periprosthetic complications were assessed. Objective and subjective indicators were compared among the image combinations. RESULTS: In the femoral component, 140â¯keV monochromatic energy images with GSI-MAR showed the lowest mean CT number, image noise, SSIM value, and streak artifacts, and the best diagnostic confidence. In the tibial component, the image noise differed significantly, but the SSIM and subjective indicators were similar among the image combinations. MAR-related new artifacts were noted in 14.7% of images, and all of them were observed in only the femoral component. CONCLUSION: GSI-MAR with higher-energy monochromatic images showed fewer metal artifacts and better visualization. We recommend 140â¯keV with GSI-MAR for improving image quality and 140â¯keV without GSI-MAR for identifying MAR-related new artifacts when evaluating TKRA.