scholarly journals A Computationally Efficient Reconstruction Algorithm for Circular Cone-Beam Computed Tomography Using Shallow Neural Networks

2020 ◽  
Vol 6 (12) ◽  
pp. 135
Author(s):  
Marinus J. Lagerwerf ◽  
Daniël M. Pelt ◽  
Willem Jan Palenstijn ◽  
Kees Joost Batenburg
Keyword(s):  
Computed Tomography ◽  
Neural Networks ◽  
Deep Neural Networks ◽  
Reconstruction Algorithm ◽  
Circular Cone ◽  
Alternative Methods ◽  
Cone Beam ◽  
Reconstruction Algorithms ◽  
Reconstruction Accuracy ◽  
Fdk Algorithm

Circular cone-beam (CCB) Computed Tomography (CT) has become an integral part of industrial quality control, materials science and medical imaging. The need to acquire and process each scan in a short time naturally leads to trade-offs between speed and reconstruction quality, creating a need for fast reconstruction algorithms capable of creating accurate reconstructions from limited data. In this paper, we introduce the Neural Network Feldkamp–Davis–Kress (NN-FDK) algorithm. This algorithm adds a machine learning component to the FDK algorithm to improve its reconstruction accuracy while maintaining its computational efficiency. Moreover, the NN-FDK algorithm is designed such that it has low training data requirements and is fast to train. This ensures that the proposed algorithm can be used to improve image quality in high-throughput CT scanning settings, where FDK is currently used to keep pace with the acquisition speed using readily available computational resources. We compare the NN-FDK algorithm to two standard CT reconstruction algorithms and to two popular deep neural networks trained to remove reconstruction artifacts from the 2D slices of an FDK reconstruction. We show that the NN-FDK reconstruction algorithm is substantially faster in computing a reconstruction than all the tested alternative methods except for the standard FDK algorithm and we show it can compute accurate CCB CT reconstructions in cases of high noise, a low number of projection angles or large cone angles. Moreover, we show that the training time of an NN-FDK network is orders of magnitude lower than the considered deep neural networks, with only a slight reduction in reconstruction accuracy.

scholarly journals Manual kidney stone size measurements in computed tomography are most accurate using multiplanar image reformatations and bone window settings

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Robert Peter Reimer ◽  
Konstantin Klein ◽  
Miriam Rinneburger ◽  
David Zopfs ◽  
Simon Lennartz ◽  
...  
Keyword(s):  
Computed Tomography ◽  
Treatment Decision ◽  
Reconstruction Algorithm ◽  
Slice Thickness ◽  
Reconstruction Algorithms ◽  
Stone Size ◽  
Bone Window ◽  
3D Printed ◽  
Measurement Strategies ◽  
Digital Caliper

AbstractComputed tomography in suspected urolithiasis provides information about the presence, location and size of stones. Particularly stone size is a key parameter in treatment decision; however, data on impact of reformatation and measurement strategies is sparse. This study aimed to investigate the influence of different image reformatations, slice thicknesses and window settings on stone size measurements. Reference stone sizes of 47 kidney stones representative for clinically encountered compositions were measured manually using a digital caliper (Man-M). Afterwards stones were placed in a 3D-printed, semi-anthropomorphic phantom, and scanned using a low dose protocol (CTDIvol 2 mGy). Images were reconstructed using hybrid-iterative and model-based iterative reconstruction algorithms (HIR, MBIR) with different slice thicknesses. Two independent readers measured largest stone diameter on axial (2 mm and 5 mm) and multiplanar reformatations (based upon 0.67 mm reconstructions) using different window settings (soft-tissue and bone). Statistics were conducted using ANOVA ± correction for multiple comparisons. Overall stone size in CT was underestimated compared to Man-M (8.8 ± 2.9 vs. 7.7 ± 2.7 mm, p < 0.05), yet closely correlated (r = 0.70). Reconstruction algorithm and slice thickness did not significantly impact measurements (p > 0.05), while image reformatations and window settings did (p < 0.05). CT measurements using multiplanar reformatation with a bone window setting showed closest agreement with Man-M (8.7 ± 3.1 vs. 8.8 ± 2.9 mm, p < 0.05, r = 0.83). Manual CT-based stone size measurements are most accurate using multiplanar image reformatation with a bone window setting, while measurements on axial planes with different slice thicknesses underestimate true stone size. Therefore, this procedure is recommended when impacting treatment decision.

Isotope computed tomography using cone-beam geometry: a comparison of two reconstruction algorithms

1987 ◽  
Vol 32 (10) ◽  
pp. 1221-1235 ◽  
Author(s):  
A Horsman ◽  
J Sutcliffe ◽  
L Burkinshaw ◽  
P Wild ◽  
J Skilling ◽  
...  
Keyword(s):  
Computed Tomography ◽  
Cone Beam ◽  
Reconstruction Algorithms ◽  
Beam Geometry

scholarly journals Combining Acceleration Techniques for Low-Dose X-Ray Cone Beam Computed Tomography Image Reconstruction

2017 ◽  
Vol 2017 ◽  
pp. 1-10
Author(s):  
Hsuan-Ming Huang ◽  
Ing-Tsung Hsiao
Keyword(s):  
Computed Tomography ◽  
Image Reconstruction ◽  
Low Dose ◽  
Computational Cost ◽  
Reconstruction Algorithm ◽  
Reconstruction Algorithms ◽  
Acceleration Techniques ◽  
Reconstruction Methods ◽  
Speed Up ◽  
Phantom Studies

Background and Objective. Over the past decade, image quality in low-dose computed tomography has been greatly improved by various compressive sensing- (CS-) based reconstruction methods. However, these methods have some disadvantages including high computational cost and slow convergence rate. Many different speed-up techniques for CS-based reconstruction algorithms have been developed. The purpose of this paper is to propose a fast reconstruction framework that combines a CS-based reconstruction algorithm with several speed-up techniques.Methods. First, total difference minimization (TDM) was implemented using the soft-threshold filtering (STF). Second, we combined TDM-STF with the ordered subsets transmission (OSTR) algorithm for accelerating the convergence. To further speed up the convergence of the proposed method, we applied the power factor and the fast iterative shrinkage thresholding algorithm to OSTR and TDM-STF, respectively.Results. Results obtained from simulation and phantom studies showed that many speed-up techniques could be combined to greatly improve the convergence speed of a CS-based reconstruction algorithm. More importantly, the increased computation time (≤10%) was minor as compared to the acceleration provided by the proposed method.Conclusions. In this paper, we have presented a CS-based reconstruction framework that combines several acceleration techniques. Both simulation and phantom studies provide evidence that the proposed method has the potential to satisfy the requirement of fast image reconstruction in practical CT.

Automated vessel segmentation in lung CT and CTA images via deep neural networks

2021 ◽  
pp. 1-15
Author(s):  
Wenjun Tan ◽  
Luyu Zhou ◽  
Xiaoshuo Li ◽  
Xiaoyu Yang ◽  
Yufei Chen ◽  
...  
Keyword(s):  
Neural Network ◽  
Computed Tomography ◽  
Neural Networks ◽  
Deep Neural Network ◽  
Deep Neural Networks ◽  
Vessel Segmentation ◽  
Multi Scale ◽  
Vascular Segmentation ◽  
Study Results ◽  
Lung Ct

BACKGROUND: The distribution of pulmonary vessels in computed tomography (CT) and computed tomography angiography (CTA) images of lung is important for diagnosing disease, formulating surgical plans and pulmonary research. PURPOSE: Based on the pulmonary vascular segmentation task of International Symposium on Image Computing and Digital Medicine 2020 challenge, this paper reviews 12 different pulmonary vascular segmentation algorithms of lung CT and CTA images and then objectively evaluates and compares their performances. METHODS: First, we present the annotated reference dataset of lung CT and CTA images. A subset of the dataset consisting 7,307 slices for training and 3,888 slices for testing was made available for participants. Second, by analyzing the performance comparison of different convolutional neural networks from 12 different institutions for pulmonary vascular segmentation, the reasons for some defects and improvements are summarized. The models are mainly based on U-Net, Attention, GAN, and multi-scale fusion network. The performance is measured in terms of Dice coefficient, over segmentation ratio and under segmentation rate. Finally, we discuss several proposed methods to improve the pulmonary vessel segmentation results using deep neural networks. RESULTS: By comparing with the annotated ground truth from both lung CT and CTA images, most of 12 deep neural network algorithms do an admirable job in pulmonary vascular extraction and segmentation with the dice coefficients ranging from 0.70 to 0.85. The dice coefficients for the top three algorithms are about 0.80. CONCLUSIONS: Study results show that integrating methods that consider spatial information, fuse multi-scale feature map, or have an excellent post-processing to deep neural network training and optimization process are significant for further improving the accuracy of pulmonary vascular segmentation.

Peripheral bronchial luminal conspicuity on dynamic-ventilation computed tomography: association with radiation doses and temporal resolution by using an ex vivo porcine lung phantom

2020 ◽  
Vol 61 (12) ◽  
pp. 1608-1617
Author(s):  
Yukihiro Nagatani ◽  
Makoto Yoshigoe ◽  
Shinsuke Tsukagoshi ◽  
Noritoshi Ushio ◽  
Kohei Ohashi ◽  
...  
Keyword(s):  
Computed Tomography ◽  
Temporal Resolution ◽  
Ex Vivo ◽  
Reconstruction Algorithm ◽  
Radiation Doses ◽  
Reconstruction Algorithms ◽  
Movement Velocity ◽  
Full Reconstruction ◽  
Luminal Area ◽  
Lung Phantom

Background It is still unclear which image reconstruction algorithm is appropriate for peripheral bronchial luminal conspicuity (PBLC) on dynamic-ventilation computed tomography (DVCT). Purpose To assess the influence of radiation doses and temporal resolution (TR) on the association between movement velocity (MV) and PBLC on DVCT. Material and Methods An ex vivo porcine lung phantom with simulated respiratory movement was scanned by 320-row CT at 240 mA and 10 mA. Peak and dip CT density and luminal area adjusted by values at end-inspiration (CTDpeak and CTDdip, luminal area ratio [LAR]) for PBLC and MVs were measured and visual scores (VS) were obtained at 12 measurement points on 13 frame images obtained at half and full reconstructions (TR 340 and 190 ms) during expiration. Size-specific dose estimate (SSDE) was applied to presume radiation dose. VS, CTDpeak, CTDdip, LAR, and their cross-correlation coefficients with MV (CCC) were compared among four methods with combinations of two reconstruction algorithms and two doses. Results The dose at 10 mA was presumed as 26 mA by SSDE for standard proportion adults. VS, CTDdip, CTDpeak, and LAR with half reconstruction at 10 mA (2.52 ± 0.59, 1.016 ± 0.221, 0.948 ± 0.103, and 0.990 ± 0.527) were similar to those at 240 mA except for VS, and different from those with full reconstruction at both doses (2.24 ± 0.85, 0.830 ± 0.209, 0.986 ± 0.065, and 1.012 ± 0.438 at 240 mA) ( P < 0.05). CCC for CTDdip with half reconstruction (–0.024 ± 0.552) at 10 mA was higher compared with full reconstruction (–0.503 ± 0.291) ( P < 0.05). Conclusion PBLC with half reconstruction at 10 mA was comparable to that at 240 mA and better than those with full reconstruction on DVCT.

ECG gated circular cone-beam multi-cycle short-scan reconstruction algorithm

2007 ◽  
Author(s):  
Udo van Stevendaal ◽  
Peter Koken ◽  
Philipp G. C. Begemann ◽  
Ralf Koester ◽  
Gerhard Adam ◽  
...  
Keyword(s):  
Reconstruction Algorithm ◽  
Circular Cone ◽  
Cone Beam
Sign in / Sign up

Export Citation Format

Share Document