Patient-specific quantification of image quality: An automated technique for measuring the distribution of organ Hounsfield units in clinical chest CT images

2017 ◽  
Vol 44 (9) ◽  
pp. 4736-4746 ◽  
Author(s):  
Ehsan Abadi ◽  
Jeremiah Sanders ◽  
Ehsan Samei
Keyword(s):  
Image Quality ◽  
Ct Images ◽  
Chest Ct ◽  
Patient Specific ◽  
Hounsfield Units ◽  
Automated Technique

scholarly journals Image quality assessment of pediatric chest and abdomen CT by deep learning reconstruction

2021 ◽  
Author(s):  
Haesung Yoon ◽  
Jisoo Kim ◽  
Hyun Ji Lim ◽  
Mi-Jung Lee
Keyword(s):  
Deep Learning ◽  
Image Quality ◽  
Noise Reduction ◽  
Iterative Reconstruction ◽  
Ct Images ◽  
Chest Ct ◽  
Abdomen Ct ◽  
Contrast Enhanced ◽  
Pediatric Chest ◽  
Body Ct

Abstract Background Efforts to reduce the radiation dose have continued steadily, with new reconstruction techniques. Recently, image denoising algorithms using artificial neural networks, termed deep learning reconstruction (DLR), have been applied to CT image reconstruction to overcome the drawbacks of iterative reconstruction (IR). The purpose of our study was to compare objective and subjective image quality of DLR and IR on pediatric abdomen and chest CT images.Methods This retrospective study included pediatric body CT images from February 2020 to October 2020, performed on 51 patients (34 boys and 17 girls; age 1–18 years). Non-contrast chest CT (n = 16), contrast-enhanced chest CT (n = 12), and contrast-enhanced abdomen CT (n = 23) images were included. Standard 50% adaptive statistical iterative reconstruction V (ASIR-V) images were compared to images with 100% ASIR-V and DLR at medium and high strengths. Attenuation, noise, contrast to noise ratio (CNR), and signal to noise (SNR) measurements were performed. Overall image quality, artifacts, and noise were subjectively assessed by two radiologists using a four-point scale (superior, average, suboptimal, and unacceptable). Quantitative and qualitative parameters were compared using repeated measures analysis of variance (ANOVA) with Bonferroni correction and Wilcoxon signed-rank tests.Results DLR had better CNR and SNR than 50% ASIR-V in both pediatric chest and abdomen CT images. When compared with 50% ASIR-V, high strength DLR was associated with noise reduction in non-contrast chest CT (33.0%), contrast-enhanced chest CT (39.6%), and contrast-enhanced abdomen CT (38.7%) with increases in CNR at 149.1%, 105.8% and 53.1% respectively. The subjective assessment of overall image quality and noise was also better on DLR images (p < 0.001). However, there was no significant difference in artifacts between reconstruction methods.Conclusion Compared with 50% ASIR-V, DLR improved pediatric body CT images with significant noise reduction. However, artifacts were not improved by DLR, regardless of strength.

The image quality of deep-learning image reconstruction of chest CT images on a mediastinal window setting

2021 ◽  
Vol 76 (2) ◽  
pp. 155.e15-155.e23
Author(s):  
A. Hata ◽  
M. Yanagawa ◽  
Y. Yoshida ◽  
T. Miyata ◽  
N. Kikuchi ◽  
...  
Keyword(s):  
Deep Learning ◽  
Image Reconstruction ◽  
Image Quality ◽  
Ct Images ◽  
Chest Ct

scholarly journals Image quality assessment of pediatric chest and abdomen CT by deep learning reconstruction

2021 ◽  
Vol 21 (1) ◽  
Author(s):  
Haesung Yoon ◽  
Jisoo Kim ◽  
Hyun Ji Lim ◽  
Mi-Jung Lee
Keyword(s):  
Deep Learning ◽  
Image Quality ◽  
Noise Reduction ◽  
Iterative Reconstruction ◽  
Ct Images ◽  
Chest Ct ◽  
Abdomen Ct ◽  
Contrast Enhanced ◽  
Pediatric Chest ◽  
Body Ct

Abstract Background Efforts to reduce the radiation dose have continued steadily, with new reconstruction techniques. Recently, image denoising algorithms using artificial neural networks, termed deep learning reconstruction (DLR), have been applied to CT image reconstruction to overcome the drawbacks of iterative reconstruction (IR). The purpose of our study was to compare the objective and subjective image quality of DLR and IR on pediatric abdomen and chest CT images. Methods This retrospective study included pediatric body CT images from February 2020 to October 2020, performed on 51 patients (34 boys and 17 girls; age 1–18 years). Non-contrast chest CT (n = 16), contrast-enhanced chest CT (n = 12), and contrast-enhanced abdomen CT (n = 23) images were included. Standard 50% adaptive statistical iterative reconstruction V (ASIR-V) images were compared to images with 100% ASIR-V and DLR at medium and high strengths. Attenuation, noise, contrast to noise ratio (CNR), and signal to noise (SNR) measurements were performed. Overall image quality, artifacts, and noise were subjectively assessed by two radiologists using a four-point scale (superior, average, suboptimal, and unacceptable). A phantom scan was performed including the dose range of the clinical images used in our study, and the noise power spectrum (NPS) was calculated. Quantitative and qualitative parameters were compared using repeated-measures analysis of variance (ANOVA) with Bonferroni correction and Wilcoxon signed-rank tests. Results DLR had better CNR and SNR than 50% ASIR-V in both pediatric chest and abdomen CT images. When compared with 50% ASIR-V, high strength DLR was associated with noise reduction in non-contrast chest CT (33.0%), contrast-enhanced chest CT (39.6%), and contrast-enhanced abdomen CT (38.7%) with increases in CNR at 149.1%, 105.8%, and 53.1% respectively. The subjective assessment of overall image quality and the noise was also better on DLR images (p < 0.001). However, there was no significant difference in artifacts between reconstruction methods. From NPS analysis, DLR methods showed a pattern of reducing the magnitude of noise while maintaining the texture. Conclusion Compared with 50% ASIR-V, DLR improved pediatric body CT images with significant noise reduction. However, artifacts were not improved by DLR, regardless of strength.

scholarly journals P.169 Reduced radiation CT imaging for augmented reality spinal surgery applications

2021 ◽  
Vol 48 (s3) ◽  
pp. S68-S68
Author(s):  
M de Lotbiniere-Bassett ◽  
E Schonfeld ◽  
T Jansen ◽  
D Anthony ◽  
A Veeravagu
Keyword(s):  
Image Quality ◽  
Augmented Reality ◽  
Radiation Dose ◽  
Spinal Surgery ◽  
Ct Images ◽  
Patient Specific ◽  
Future Research ◽  
Pedicle Screw Placement ◽  
Quality Threshold ◽  
Ct Image Quality

Background: There is growing evidence for the use of augmented reality (AR) in pedicle screw placement in spinal surgery to increase surgical accuracy, improve clinical outcomes and reduce the radiation exposure required for intraoperative navigation. Auto-segmentation is the cornerstone of AR applications because it correlates patient-specific anatomy to structures segmented from preoperative computed tomography (pCT) images. These AR techniques allow for a reduction in the radiation dose required to acquire CT images while maintaining accurate segmentation. Methods: In this study, we methodically increase the noise that is introduced into CT images to determine the image quality threshold that is required for auto-segmentation on pCT. We then enhance the images with denoising algorithms to evaluate the effect on the segmentation. Results: The pCT radiation dose is decreased to below the current lowest clinical threshold and the resulting images still produce segmentations that are appropriate for input into AR applications. The application of denoising algorithms to the images resulted in increased artifacts and decreased bone density. Conclusions: The CT image quality that is required for successful AR auto-segmentation is lower than that which is currently employed in spine surgery. Future research is required to identify the specific, clinically relevant radiation dose thresholds.

scholarly journals Assessment of 11 Available Materials With Custom Three-Dimensional-Printing Patterns for the Simulation of Muscle, Fat, and Lung Hounsfield Units in Patient-Specific Phantoms

2017 ◽  
Vol 1 (1) ◽  
Author(s):  
Nikiforos Okkalidis ◽  
Chrysoula Chatzigeorgiou ◽  
Demetrios Okkalides
Keyword(s):  
Fused Deposition Modeling ◽  
Three Dimensional ◽  
Region Of Interest ◽  
Ct Images ◽  
Effective Density ◽  
Patient Specific ◽  
Hounsfield Units ◽  
Fused Deposition ◽  
3D Printers ◽  
Deposition Modeling

A couple of fused deposition modeling (FDM) three-dimensional (3D) printers using variable infill density patterns were employed to simulate human muscle, fat, and lung tissue as it is represented by Hounsfield units (HUs) in computer tomography (CT) scans. Eleven different commercial plastic filaments were assessed by measuring their mean HU on CT images of small cubes printed with different patterns. The HU values were proportional to the mean effective density of the cubes. Polylactic acid (PLA) filaments were chosen. They had good printing characteristics and acceptable HU. Such filaments obtained from two different vendors were then tested by printing two sets of cubes comprising 10 and 6 cubes with 100% to 20% and 100% to 50% infill densities, respectively. They were printed with different printing patterns named “Regular” and “Bricks,” respectively. It was found that the HU values measured on the CT images of the 3D-printed cubes were proportional to the infill density with slight differences between vendors and printers. The Regular pattern with infill densities of about 30%, 90%, and 100% were found to produce HUs equivalent to lung, fat, and muscle. This was confirmed with histograms of the respective region of interest (ROI). The assessment of popular 3D-printing materials resulted in the choice of PLA, which together with the proposed technique was found suitable for the adequate simulation of the muscle, fat, and lung HU in printed patient-specific phantoms.

scholarly journals Scan Quality and Entrance Skin Dose in Thoracic CT: A Comparison between Bismuth Breast Shield and Posteriorly Centered Partial CT Scans

2013 ◽  
Vol 2013 ◽  
pp. 1-6 ◽  
Author(s):  
Rafel Tappouni ◽  
Bradley Mathers
Keyword(s):  
Image Quality ◽  
Anterior Chest Wall ◽  
Chest Ct ◽  
Ct Scans ◽  
Skin Dose ◽  
Hounsfield Units ◽  
Anthropomorphic Phantom ◽  
Thoracic Ct ◽  
Entrance Skin Dose ◽  
Effective Doses

Objectives. To compare the effectiveness of the bismuth breast shield and partial CT scan in reducing entrance skin dose and to evaluate the effect of the breast shield on image quality (IQ). Methods. Nanodots were placed on an adult anthropomorphic phantom. Standard chest CT, CT with shield, and partial CT were performed. Nanodot readings and effective doses were recorded. 50 patients with chest CTs obtained both with and without breast shields were reviewed. IQ was evaluated by two radiologists and by measuring Hounsfield units (HUs) and standard deviation (SD) of HU in anterior subcutaneous region. Results. Breast shield and the partial CT scans reduced radiation to the anterior chest by 38% and 16%, respectively. Partial CT increased dose to the posterior chest by 37% and effective dose by 8%. Change in IQ in shield CT was observed in the anterior chest wall. Significant change in IQ was observed in 5/50 cases. The shield caused an increase of 20 HU () and a 1.86 reduction in SD of HU () in the anterior compared to posterior subcutaneous regions. Summary. Bismuth breast shield is more effective than the partial CT in reducing entrance skin dose while maintaining image quality.

COVID-19 pneumonia diagnosis using a simple 2D deep learning framework with a single chest CT image (Preprint)

2020 ◽  
Author(s):  
Jinseok Lee
Keyword(s):  
Deep Learning ◽  
Diagnostic Performance ◽  
Ct Images ◽  
Chest Ct ◽  
University Hospital ◽  
Detection Accuracy ◽  
Ct Image ◽  
Test Dataset ◽  
Learning Framework ◽  
Testing Dataset

BACKGROUND The coronavirus disease (COVID-19) has explosively spread worldwide since the beginning of 2020. According to a multinational consensus statement from the Fleischner Society, computed tomography (CT) can be used as a relevant screening tool owing to its higher sensitivity for detecting early pneumonic changes. However, physicians are extremely busy fighting COVID-19 in this era of worldwide crisis. Thus, it is crucial to accelerate the development of an artificial intelligence (AI) diagnostic tool to support physicians. OBJECTIVE We aimed to quickly develop an AI technique to diagnose COVID-19 pneumonia and differentiate it from non-COVID pneumonia and non-pneumonia diseases on CT. METHODS A simple 2D deep learning framework, named fast-track COVID-19 classification network (FCONet), was developed to diagnose COVID-19 pneumonia based on a single chest CT image. FCONet was developed by transfer learning, using one of the four state-of-art pre-trained deep learning models (VGG16, ResNet50, InceptionV3, or Xception) as a backbone. For training and testing of FCONet, we collected 3,993 chest CT images of patients with COVID-19 pneumonia, other pneumonia, and non-pneumonia diseases from Wonkwang University Hospital, Chonnam National University Hospital, and the Italian Society of Medical and Interventional Radiology public database. These CT images were split into a training and a testing set at a ratio of 8:2. For the test dataset, the diagnostic performance to diagnose COVID-19 pneumonia was compared among the four pre-trained FCONet models. In addition, we tested the FCONet models on an additional external testing dataset extracted from the embedded low-quality chest CT images of COVID-19 pneumonia in recently published papers. RESULTS Of the four pre-trained models of FCONet, the ResNet50 showed excellent diagnostic performance (sensitivity 99.58%, specificity 100%, and accuracy 99.87%) and outperformed the other three pre-trained models in testing dataset. In additional external test dataset using low-quality CT images, the detection accuracy of the ResNet50 model was the highest (96.97%), followed by Xception, InceptionV3, and VGG16 (90.71%, 89.38%, and 87.12%, respectively). CONCLUSIONS The FCONet, a simple 2D deep learning framework based on a single chest CT image, provides excellent diagnostic performance in detecting COVID-19 pneumonia. Based on our testing dataset, the ResNet50-based FCONet might be the best model, as it outperformed other FCONet models based on VGG16, Xception, and InceptionV3.

Sign in / Sign up

Export Citation Format

Share Document