Chest Computed Tomography Image Quality Research Articles

Deep learning-driven multi-view multi-task image quality assessment method for chest CT image

BackgroundChest computed tomography (CT) image quality impacts radiologists’ diagnoses. Pre-diagnostic image quality assessment is essential but labor-intensive and may have human limitations (fatigue, perceptual biases, and cognitive biases). This study aims to develop and validate a deep learning (DL)-driven multi-view multi-task image quality assessment (M2\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$^2$$\\end{document}IQA) method for assessing the quality of chest CT images in patients, to determine if they are suitable for assessing the patient’s physical condition.MethodsThis retrospective study utilizes and analyzes chest CT images from 327 patients. Among them, 1613 images from 286 patients are used for model training and validation, while the remaining 41 patients are reserved as an additional test set for conducting ablation studies, comparative studies, and observer studies. The M2\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$^2$$\\end{document}IQA method is driven by DL technology and employs a multi-view fusion strategy, which incorporates three scanning planes (coronal, axial, and sagittal). It assesses image quality for multiple tasks, including inspiration evaluation, position evaluation, radiation protection evaluation, and artifact evaluation. Four algorithms (pixel threshold, neural statistics, region measurement, and distance measurement) have been proposed, each tailored for specific evaluation tasks, with the aim of optimizing the evaluation performance of the M2\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$^2$$\\end{document}IQA method.ResultsIn the additional test set, the M2\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$^2$$\\end{document}IQA method achieved 87% precision, 93% sensitivity, 69% specificity, and a 0.90 F1-score. Extensive ablation and comparative studies have demonstrated the effectiveness of the proposed algorithms and the generalization performance of the proposed method across various assessment tasks.ConclusionThis study develops and validates a DL-driven M2\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$^2$$\\end{document}IQA method, complemented by four proposed algorithms. It holds great promise in automating the assessment of chest CT image quality. The performance of this method, as well as the effectiveness of the four algorithms, is demonstrated on an additional test set.

Automatic chest computed tomography image noise quantification using deep learning

PurposeThis study aimed to develop a deep learning (DL) method for noise quantification for clinical chest computed tomography (CT) images without the need for repeated scanning or homogeneous tissue regions. MethodsA comprehensive phantom CT dataset (three dose levels, six reconstruction methods, amounting to 9240 slices) was acquired and used to train a convolutional neural network (CNN) to output an estimate of local image noise standard deviations (SD) from a single CT scan input. The CNN model consisting of seven convolutional layers was trained on the phantom image dataset representing a range of scan parameters and was tested with phantom images acquired in a variety of different scan conditions, as well as publicly available chest CT images to produce clinical noise SD maps. ResultsNoise SD maps predicted by the CNN agreed well with the ground truth both visually and numerically in the phantom dataset (errors of < 5 HU for most scan parameter combinations). In addition, the noise SD estimates obtained from clinical chest CT images were similar to running-average based reference estimates in areas without prominent tissue interfaces. ConclusionsPredicting local noise magnitudes without the need for repeated scans is feasible using DL. Our implementation trained with phantom data was successfully applied to open-source clinical data with heterogeneous tissue borders and textures. We suggest that automatic DL noise mapping from clinical patient images could be used as a tool for objective CT image quality estimation and protocol optimization.

Impact of tin filter on the image quality of ultra-low dose chest CT: A phantom study on three CT systems

PurposeThe purpose of this study was to assess the impact of a tin filter on the image quality of ultra-low dose (ULD) chest computed tomography (CT) on three different CT systems. Materials and methodsAn image quality phantom was scanned on three CT systems including two split-filter dual-energy CT (SFCT-1 and SFCT-2) scanners and one dual-source CT scanner (DSCT). Acquisitions were performed with a volume CT dose index (CTDIvol) of 0.4 mGy, first at 100 kVp without tin filter (Sn), and second, at Sn100/Sn140 kVp, Sn100/Sn110/Sn120/Sn130/Sn140/Sn150 kVp and Sn100/Sn150 kVp for SFCT-1, SFCT-2 and DSCT respectively. Noise-power-spectrum and task-based transfer function were computed. The detectability index (d’) was computed to model the detection of two chest lesions. ResultsFor DSCT and SFCT-1, noise magnitude values were higher with 100kVp than with Sn100 kVp and with Sn140 kVp or Sn150 kVp than with Sn100 kVp. For SFCT-2, noise magnitude increased from Sn110 kVp to Sn150 kVp and was higher at Sn100 kVp than at Sn110 kVp. For most kVp with the tin filter, the noise amplitude values were lower than those obtained at 100 kVp. For each CT system, noise texture and spatial resolution values were similar with 100 kVp and with all kVp used with a tin filter. For all simulated chest lesions, the highest d’ values were obtained at Sn100 kVp for SFCT-1 and DSCT and at Sn110 kVp for SFCT-2. ConclusionFor ULD chest CT protocols, the lowest noise magnitude and highest detectability values for simulated chest lesions are obtained with Sn100 kVp for the SFCT-1 and DSCT CT systems and at Sn110 kVp for SFCT-2.

Feasibility of Deep Learning-Based Noise and Artifact Reduction in Coronal Reformation of Contrast-Enhanced Chest Computed Tomography.

This study aimed to evaluate the feasibility of a deep learning method for imaging artifact and noise reduction in coronal reformation of contrast-enhanced chest computed tomography (CT). A total of 19,052 coronal reformatted chest CT images of 110 CT image sets (55 pairs of concordant 16- and 320-row CT image sets) were included and used to train a deep learning algorithm for artifact and noise correction. For internal validation, 4093 coronal reformatted CT images of 25 patients from 16-row CT images underwent correction processing. For external validation, chest CT images of 30 patients (1028 coronal reformatted CT images), acquired in other institutions using different scanners, were subjected to correction processing. For both validations, image quality was compared between original ("CT origin ") and deep learning-based corrected ("CT correct ") CT images. Quantitative analysis for stair-step artifact (coefficient of variance of CT density on coronal reformation), image noise, signal-to-noise ratio, and contrast-to-noise ratio were evaluated. Subjective image quality scores were assigned for image contrast, artifact, and conspicuity of major structures. CT correct showed significantly reduced stair-step artifact (mean coefficient of variance: CT origin 7.35 ± 2.0 vs CT correct 5.17 ± 2.4, P < 0.001) and image noise and improved signal-to-noise ratio and contrast-to-noise ratio in the aorta, pulmonary artery, and liver, compared with those of CT origin ( P < 0.01). On subjective analysis, CT correct had higher image contrast, lower artifact, and better conspicuity than CT origin . Most results of the external validation were consistent with those obtained from the internal validation, except for those concerning the pulmonary artery. Deep learning-based artifact correction significantly improved the image quality of coronal reformation chest CT by reducing image noise and artifacts.

Difference in Computed Tomography Image Quality between Central Vein and Peripheral Vein Enhancement in Treatment Naive Esophageal Cancer Patients.

Simple SummaryA chest CT via central vein enhancement not only eliminates peripheral vein regurgitation but also provides better image quality that facilitates precise clinical staging. A chest CT via central vein enhancement may be considered after tissue proof in order to better discriminate disease severity.The differences in chest computed tomography (CT) image quality may affect the tumor stage. The aim of this study was to compare the image quality and accuracy of chest CT via central vein and peripheral vein enhancement. Fifty consecutive patients were enrolled from a tertiary medical center in Taiwan from May 2016 to March 2019. All the patients received a chest CT via central vein enhancement prior to neoadjuvant concurrent chemoradiation in order to compare the chest CT that was obtained via the peripheral vein. In addition, blind independent central reviews of chest CT via central vein and peripheral vein enhancement were conducted. For T and N stage, chest CT via central vein enhancement had a greater consistency with endoscopic ultrasonography and positron-emission tomography-computed tomography findings (kappa coefficients 0.4471 and 0.5564, respectively). In addition, chest CT via central vein enhancement also showed excellent agreement in the blind independent central review (kappa coefficient 0.9157). The changes in the T and N stage resulted in stage migration in 16 patients. Chest CT via central vein enhancement eliminated peripheral vein regurgitation and also provided more precise clinical staging. This study is registered under the registered NCT number 02887261.

Usefulness of dictionary learning-based processing for improving image quality of sub-millisievert low-dose chest CT: initial experience.

To develop a dictionary learning (DL)-based processing technique for improving the image quality of sub-millisievert chest computed tomography (CT). Standard-dose and sub-millisievert chest CT were acquired in 12 patients. Dictionaries including standard- and low-dose image patches were generated from the CT datasets. For each patient, DL-based processing was performed for low-dose CT using the dictionaries generated from the remaining 11 patients. This procedure was repeated for all 12 patients. Image quality of normal thoracic structures on the processed sub-millisievert CT images was assessed with a 5-point scale (5 = excellent, 1 = very poor). Lung lesion conspicuity was also assessed on a 5-point scale. Image noise on sub-millisievert CT was significantly decreased with DL-based image processing (48.5 ± 13.7 HU vs 20.4 ± 7.9 HU, p = 0.0005). Image quality of lung structures was significantly improved with DL-based method (middle level of lung, 2.25 ± 0.75 vs 2.92 ± 0.79, p = 0.0078). Lung lesion conspicuity was also significantly improved with DL-based technique (solid nodules, 3.4 ± 0.6 vs 2.7 ± 0.6, p = 0.0273). Image quality and lesion conspicuity on sub-millisievert chest CT images may be improved by DL-based post-processing.

Image quality evaluation for CARE kV technique combined with iterative reconstruction for chest computed tomography scanning.

Background:To investigate the radiation dose and image quality for iterative reconstruction combined with the CARE kV technique in chest computed tomography (CT) scanning for physical examination.Methods:A total of 130 patients who underwent chest CT scanning were randomly chosen and the quality reference value was set as 80 mAs. The scanning scheme was set and the patients were randomly divided into groups according to the scanning scheme. Sixty patients underwent a chest scan with 100 kV using the CARE kV technique and SAFIRE reconstruction (value=3) (experimental group) and the other 70 patients underwent chest scanning with 120 kV (control group). The mean CT value, image noise (SD), and signal-to-noise ratio (SNR) of the apex of the lung, the level of the descending aorta bifurcation of the trachea, and the middle area of the left atrium were measured. The image quality was assessed on a 5-point scale by two radiologists and results of the two groups were compared. The CT dose index of the volume (CTDIvol), dose length product (DLP), and effective dose (ED) were compared.Results:All the images for both groups satisfied the diagnosis requirement. There was no statistical difference in the image quality between the two methods (P > 0.05). The mean CT value of the apex of the lung, the level of the descending aorta bifurcation of the trachea, and the middle area of the left atrium were not significantly different for both groups (P > 0.05), while the image noise (SD) and the signal-to-noise ratio (SNR) of the apex of the lung, the level of the descending aorta bifurcation of the trachea, and the middle area of the left atrium were statistically different for both groups (P < 0.05). The CTDIvol was 3.29 ± 1.17 mGy for the experimental group and 5.30 ± 1.53 mGy for the control group. The DLP was 114.9 ± 43.73 mGy cm for the low-dose group and 167.6 ± 44.59 mGy cm for the control group. The ED was 1.61 ± 0.61 mSv for the low-dose group and 2.35 ± 0.62 mSv for the control group (P < 0.05).Conclusion:The CARE kV technique combined with iterative reconstruction for chest CT scanning for physical examination could reduce the radiation dosage and improve CT image quality, which has a potential clinical value for imaging the thorax.

CT chest and gantry rotation time: does the rotation time influence image quality?

Computed tomography (CT) gantry rotation time is one factor influencing image quality. Until now, there has been no report investigating the influence of gantry rotation time on chest CT image quality. To investigate the influence of faster gantry rotation time on image quality and subjective and objective image parameters in chest CT imaging. Chest CT scans from 160 patients were examined in this study. All scans were performed using a single-source mode (collimation, 128 × 0.6 mm; pitch, 1.2) on a dual-source CT scanner. Only gantry rotation time was modified, while other CT parameters were kept stable for each scan (120 kV/110 reference mAs). Patients were divided into four groups based on rotation time: group 1, 1 s/ rotation (rot); group 2, 0.5 s/rot; group 3, 0.33 s/rot; group 4, 0.28 s/rot. Two blinded radiologists subjectively compared CT image quality, noise, and artifacts, as well as radiation exposure, from all groups. For objective comparison, all image datasets were analyzed by a radiologist with 5 years of experience concerning objective measurements as well as signal-to-noise ratio (SNR). We found that faster gantry rotation times (0.28 s/rot and 0.33 s/rot) resulted in more streak artifacts, image noise, and decreased image quality. However, there was no significant difference in radiation exposure between faster and slower rotation times (P > 0.7). Faster CT gantry rotation reduces scan time and motion artifacts. However, accelerating rotation time increases image noise and streak artifacts. Therefore, a slower CT gantry rotation speed is still recommended for higher image quality in some cases.

Clinical impact of model-based type iterative reconstruction with fast reconstruction time on image quality of low-dose screening chest CT.

Model-based type iterative reconstruction algorithms with fast reconstruction times are now available. The clinical feasibility of their reconstruction has not been evaluated adequately. To investigate the effects of model-based type iterative reconstruction, i.e. iterative model reconstruction (IMR), with fast reconstruction time on the qualitative and quantitative image quality at low-dose chest computed tomography (CT). Thirty-one patients undergoing low-dose screening chest CT were enrolled. Images were reconstructed using filtered back projection (FBP), hybrid iterative reconstruction (HIR), and IMR algorithms. The CT attenuation and image noise for all reconstructions were calculated at the lung apex, middle, and base. Using a 4-point scale, two reviewers visually evaluated the image quality with respect to vessel sharpness, streak artifact, the mediastinum, and the overall image quality of each reconstruction method. The mean estimated effective dose was 1.0 ± 0.3 mSv. There was no significant difference in the CT attenuation among the three reconstructions. The mean image noise of FBP, HIR, and IMR images was 124.3 ± 57.3, 34.8 ± 10.2, and 22.9 ± 5.8 HU, respectively. There were significant differences for all comparison combinations among the three methods (P < 0.01). The best subjective overall image quality for the lung and mediastinum was obtained with IMR (P < 0.01). The reconstruction time for IMR was within 3 min in all cases. At low-dose chest CT, IMR can improve the qualitative and quantitative visualization of both lung and mediastinal structures especially in the lung apex at a clinically acceptable reconstruction time. Its application may improve diagnostic performance.

Adaptive Iterative Dose Reduction Using Three Dimensional Processing (AIDR3D) improves chest CT image quality and reduces radiation exposure.

ObjectiveTo assess the advantages of Adaptive Iterative Dose Reduction using Three Dimensional Processing (AIDR3D) for image quality improvement and dose reduction for chest computed tomography (CT).MethodsInstitutional Review Boards approved this study and informed consent was obtained. Eighty-eight subjects underwent chest CT at five institutions using identical scanners and protocols. During a single visit, each subject was scanned using different tube currents: 240, 120, and 60 mA. Scan data were converted to images using AIDR3D and a conventional reconstruction mode (without AIDR3D). Using a 5-point scale from 1 (non-diagnostic) to 5 (excellent), three blinded observers independently evaluated image quality for three lung zones, four patterns of lung disease (nodule/mass, emphysema, bronchiolitis, and diffuse lung disease), and three mediastinal measurements (small structure visibility, streak artifacts, and shoulder artifacts). Differences in these scores were assessed by Scheffe's test.ResultsAt each tube current, scans using AIDR3D had higher scores than those without AIDR3D, which were significant for lung zones (p<0.0001) and all mediastinal measurements (p<0.01). For lung diseases, significant improvements with AIDR3D were frequently observed at 120 and 60 mA. Scans with AIDR3D at 120 mA had significantly higher scores than those without AIDR3D at 240 mA for lung zones and mediastinal streak artifacts (p<0.0001), and slightly higher or equal scores for all other measurements. Scans with AIDR3D at 60 mA were also judged superior or equivalent to those without AIDR3D at 120 mA.ConclusionFor chest CT, AIDR3D provides better image quality and can reduce radiation exposure by 50%.

Reduced computed tomography radiation dose in HIV-related pneumonia: effect on diagnostic image quality

The objective of this study was to compare the diagnostic quality of low-dose computed tomography (CT) with that of standard-dose chest CT in the diagnosis of infectious lung diseases. Thirty chest CT scans [high-resolution computed tomography (HRCT), 15; spiral CT, 15] were performed in HIV-positive patients with an infectious lung disease. Two additional slices were obtained at two lower exposures (HRCT, 120 kV/70 mAs and 120 kV/50 mAs; spiral CT, 100 kV/56 mAs and 100 kV/40 mAs) after chest routine CT. Observers compared the quality of the images obtained at different parameters and image noise. Objective evaluation of image noise was also made. Diagnostic image quality was excellent in 93% of the low-dose HRCT scans and in 86% of the low-dose spiral CT scans, rates that are always acceptable in any case. Significant differences were found in noise levels between the low-dose and reference scans; however, artifacts did not compromise detection of abnormalities. In HRCT, a mean reduction of 77% from the standard technique to the low-dose scan is possible in total and in lung effective doses. In spiral CT, this reduction is lower at 71%. These values can reach a further reduction with ultra-low-dose imaging (84% in HRCT and 80% in spiral CT). Chest CT image quality appears to be adequate to evaluate pulmonary infectious diseases, even with an effective reduction in radiation dose. Standard-dose CT with a higher patient effective dose may be appropriate for selected cases.