A semi-supervised network-based tissue-aware contrast enhancement method for CT images / 南方医科大学学报

OBJECTIVE@#To propose a tissue- aware contrast enhancement network (T- ACEnet) for CT image enhancement and validate its accuracy in CT image organ segmentation tasks.@*METHODS@#The original CT images were mapped to generate low dynamic grayscale images with lung and soft tissue window contrasts, and the supervised sub-network learned to recognize the optimal window width and level setting of the lung and abdominal soft tissues via the lung mask. The self-supervised sub-network then used the extreme value suppression loss function to preserve more organ edge structure information. The images generated by the T-ACEnet were fed into the segmentation network to segment multiple abdominal organs.@*RESULTS@#The images obtained by T-ACEnet were capable of providing more window setting information in a single image, which allowed the physicians to conduct preliminary screening of the lesions. Compared with the suboptimal methods, T-ACE images achieved improvements by 0.51, 0.26, 0.10, and 14.14 in SSIM, QABF, VIFF, and PSNR metrics, respectively, with a reduced MSE by an order of magnitude. When T-ACE images were used as input for segmentation networks, the organ segmentation accuracy could be effectively improved without changing the model as compared with the original CT images. All the 5 segmentation quantitative indices were improved, with the maximum improvement of 4.16%.@*CONCLUSION@#The T-ACEnet can perceptually improve the contrast of organ tissues and provide more comprehensive and continuous diagnostic information, and the T-ACE images generated using this method can significantly improve the performance of organ segmentation tasks.

A semi-supervised material quantitative intelligent imaging algorithm for spectral CT based on prior information perception learning / 南方医科大学学报

OBJECTIVE@#To propose a semi-supervised material quantitative intelligent imaging algorithm based on prior information perception learning (SLMD-Net) to improve the quality and precision of spectral CT imaging.@*METHODS@#The algorithm includes a supervised and a self- supervised submodule. In the supervised submodule, the mapping relationship between low and high signal-to-noise ratio (SNR) data was constructed through mean square error loss function learning based on a small labeled dataset. In the self- supervised sub-module, an image recovery model was utilized to construct the loss function incorporating the prior information from a large unlabeled low SNR basic material image dataset, and the total variation (TV) model was used to to characterize the prior information of the images. The two submodules were combined to form the SLMD-Net method, and pre-clinical simulation data were used to validate the feasibility and effectiveness of the algorithm.@*RESULTS@#Compared with the traditional model-driven quantitative imaging methods (FBP-DI, PWLS-PCG, and E3DTV), data-driven supervised-learning-based quantitative imaging methods (SUMD-Net and BFCNN), a material quantitative imaging method based on unsupervised learning (UNTV-Net) and semi-supervised learning-based cycle consistent generative adversarial network (Semi-CycleGAN), the proposed SLMD-Net method had better performance in both visual and quantitative assessments. For quantitative imaging of water and bone materials, the SLMD-Net method had the highest PSNR index (31.82 and 29.06), the highest FSIM index (0.95 and 0.90), and the lowest RMSE index (0.03 and 0.02), respectively) and achieved significantly higher image quality scores than the other 7 material decomposition methods (P < 0.05). The material quantitative imaging performance of SLMD-Net was close to that of the supervised network SUMD-Net trained with labeled data with a doubled size.@*CONCLUSIONS@#A small labeled dataset and a large unlabeled low SNR material image dataset can be fully used to suppress noise amplification and artifacts in basic material decomposition in spectral CT and reduce the dependence on labeled data-driven network, which considers more realistic scenario in clinics.

Establishment of a deep feature-based classification model for distinguishing benign and malignant breast tumors on full-filed digital mammography / 南方医科大学学报

OBJECTIVE@#To develop a deep features-based model to classify benign and malignant breast lesions on full- filed digital mammography.@*METHODS@#The data of full-filed digital mammography in both craniocaudal view and mediolateral oblique view from 106 patients with breast neoplasms were analyzed. Twenty-three handcrafted features (HCF) were extracted from the images of the breast tumors and a suitable feature set of HCF was selected using -test. The deep features (DF) were extracted from the 3 pre-trained deep learning models, namely AlexNet, VGG16 and GoogLeNet. With abundant breast tumor information from the craniocaudal view and mediolateral oblique view, we combined the two extracted features (DF and HCF) as the two-view features. A multi-classifier model was finally constructed based on the combined HCF and DF sets. The classification ability of different deep learning networks was evaluated.@*RESULTS@#Quantitative evaluation results showed that the proposed HCF+DF model outperformed HCF model, and AlexNet produced the best performances among the 3 deep learning models.@*CONCLUSIONS@#The proposed model that combines DF and HCF sets of breast tumors can effectively distinguish benign and malignant breast lesions on full-filed digital mammography.

Key technologies in digital breast tomosynthesis system:theory, design, and optimization / 南方医科大学学报

OBJECTIVE@#To develop a digital breast tomosynthesis (DBT) imaging system with optimizes imaging chain.@*METHODS@#Based on 3D tomography and DBT imaging scanning, we analyzed the methods for projection data correction, geometric correction, projection enhancement, filter modulation, and image reconstruction, and established a hardware testing platform. In the experiment, the standard ACR phantom and high-resolution phantom were used to evaluate the system stability and noise level. The patient projection data of commercial equipment was used to test the effect of the imaging algorithm.@*RESULTS@#In the high-resolution phantom study, the line pairs were clear without confusing artifacts in the images reconstructed with the geometric correction parameters. In ACR phantom study, the calcified foci, cysts, and fibrous structures were more clearly defined in the reconstructed images after filtering and modulation. The patient data study showed a high contrast between tissues, and the lesions were more clearly displayed in the reconstructed image.@*CONCLUSIONS@#This DBT imaging system can be used for mammary tomography with an image quality comparable to that of commercial DBT systems to facilitate imaging diagnosis of breast diseases.

Sparse-view helical CT reconstruction based on tensor total generalized variation minimization / 南方医科大学学报

OBJECTIVE@#We propose a sparse-view helical CT iterative reconstruction algorithm based on projection of convex set tensor total generalized variation minimization (TTGV-POCS) to reduce the X-ray dose of helical CT scanning.@*METHODS@#The three-dimensional volume data of helical CT reconstruction was viewed as the third-order tensor. The tensor generalized total variation (TTGV) was used to describe the structural sparsity of the three-dimensional image. The POCS iterative reconstruction framework was adopted to achieve a robust result of sparse-view helical CT reconstruction. The TTGV-POCS algorithm fully used the structural sparsity of first-order and second-order derivation and the correlation between the slices of helical CT image data to effectively suppress artifacts and noise in the image of sparse-view reconstruction and better preserve image edge information.@*RESULTS@#The experimental results of XCAT phantom and patient scan data showed that the TTGVPOCS algorithm had better performance in reducing noise, removing artifacts and maintaining edges than the existing reconstruction algorithms. Comparison of the sparse-view reconstruction results of XCAT phantom data with 144 exposure views showed that the TTGV-POCS algorithm proposed herein increased the PSNR quantitative index by 9.17%-15.24% compared with the experimental comparison algorithm; the FSIM quantitative index was increased by 1.27%-9.30%.@*CONCLUSIONS@#The TTGV-POCS algorithm can effectively improve the image quality of helical CT sparse-view reconstruction and reduce the radiation dose of helical CT examination to improve the clinical imaging diagnosis.

Design and optimization of a cone-beam CT system for extremity imaging / 南方医科大学学报

OBJECTIVE@#To establish a cone beam computed tomography (ECBCT) system for high-resolution imaging of the extremities.@*METHODS@#Based on three-dimensional X-Ray CT imaging and high-resolution flat plate detector technique, we constructed a physical model and a geometric model for ECBCT imaging, optimized the geometric calibration and image reconstruction methods, and established the scanner system. In the experiments, the pencil vase phantom, image quality (IQ) phantom and a swine feet were scanned using this imaging system to evaluate its effectiveness and stability.@*RESULTS@#On the reconstructed image of the pencil vase phantom, the edges were well preserved with geometric calibrated parameters and no aliasing artifacts were observed. The reconstructed images of the IQ phantom showed a uniform distribution of the CT number, and the noise power spectra were stable in multiple scanning under the same condition. The reconstructed images of the swine feet had clearly displayed the bones with a good resolution.@*CONCLUSIONS@#The ECBCT system can be used for highresolution imaging of the extremities to provide important imaging information to assist in the diagnosis of bone diseases.

Edge-detecting operator-based selection of Huber regularization threshold for low-dose computed tomography imaging / 南方医科大学学报

<p><b>OBJECTIVE</b>To compare two methods for threshold selection in Huber regularization for low-dose computed tomography imaging.</p><p><b>METHODS</b>Huber regularization-based iterative reconstruction (IR) approach was adopted for low-dose CT image reconstruction and the threshold of Huber regularization was selected based on global versus local edge-detecting operators.</p><p><b>RESULTS</b>The experimental results on the simulation data demonstrated that both of the two threshold selection methods in Huber regularization could yield remarkable gains in terms of noise suppression and artifact removal.</p><p><b>CONCLUSION</b>Both of the two methods for threshold selection in Huber regularization can yield high-quality images in low-dose CT image iterative reconstruction.</p>

Robust low-dose CT myocardial perfusion deconvolution via high-dimension total variation regularization / 南方医科大学学报

<p><b>OBJECTIVE</b>To develop a computed tomography myocardial perfusion (CT-MP) deconvolution algorithm by incorporating high-dimension total variation (HDTV) regularization.</p><p><b>METHODS</b>A perfusion deconvolution model was formulated for the low-dose CT-MPI data, followed by HDTV regularization to regularize the consistency of the solution by fusing the spatial correlation of the vascular structure and the temporal continuation of the blood flow signal.</p><p><b>RESULTS</b>Both qualitative and quantitative studies were conducted using XCAT and pig myocardial perfusion data to evaluate the present algorithm. The experimental results showed that this algorithm achieved hemodynamic parameter maps with better performances than the existing methods in terms of streak-artifacts suppression, noise-resolution tradeoff, and diagnosis structure preservation.</p><p><b>CONCLUSION</b>The proposed algorithm can achieve high-quality hemodynamic parameter maps in low-dose CT-MPI.</p>

Tsallis entropy-based prior for PET reconstruction / 生物医学工程学杂志

Maximum a Posteriori (MAP) method has been widely applied to the ill-posed problem of image reconstruction. The choice of prior is the crucial point on MAP methods. However, the most conventional priors will lead to a blurring of the whole image or cause ladder-like artifacts. We therefore proposed a Tsallis entropy-based prior for positron emission tomography (PET) iterative reconstruction in MAP framework. The method uses a Tsallis entropy-based prior to eliminate the uncertainty between prior information and the estimated images. We tested this method in the phantom image, compared it with the traditional prior methods. the results showed that the proposed algorithm could suppress noise and obtain better reconstructed image quality.

An improved prior image constrained compressed sensing reconstruction for low-dose computed tomography / 南方医科大学学报

Low-dose computed tomography (CT) reconstruction has become the focus of X-ray CT imaging study. In this paper, we propose an improved prior image constrained compressed sensing (PICCS) reconstruction approach. A penalized weighted least-squares approach was adopted to realize the line integral projection (sinogram) data restoration, followed by filtered back-projection (FBP) of the restored sinogram data for image reconstruction. Finally, the FBP image as the prior image was used for PICCS approach for dose reduction. Qualitative and quantitative evaluations were carried out with computer simulation. The results showed that the present approach yielded noticeable gains over the original PICCS approach for dose reduction in terms of noise-induced artifacts suppression and edge detail preservation.

Low-dose CT angiography image restoration using normal dose scan-induced non-local means algorithm / 南方医科大学学报

<p><b>OBJECTIVE</b>To minimize of the radiation dose of cardiovascular CT angiography (CTA) imaging while preserving the image quality.</p><p><b>METHODS</b>To reduce the radiation dose in CTA imaging, the normal-dose scan induced non-local means (ndiNLM) algorithm was adapted for low-mAs scanned CTA image restoration by using the previous scanned high-quality image.</p><p><b>RESULTS</b>Qualitative and quantitative evaluations were carried out on both simulated phantom and clinical CTA scans in terms of accuracy and resolution properties. Compared to the original NLM algorithm, the ndiNLM method could achieve noticeable gains in terms of noise-induced artifacts suppression and enhanced structure preservation.</p><p><b>CONCLUSION</b>The ndiNLM algorithm is a potential useful technique to reduce the radiation dose in CTA imaging.</p>