Medical CT Images Research Articles

Image denoising by transfer learning of generative adversarial network for dental CT

The successful development of the image denoising techniques for low-dose computed tomography (LDCT) was largely owing to the public-domain availability of spatially-aligned high- and low-dose CT image pairs. Even though low-dose CT scans are also highly desired in dental imaging, public-domain databases of dental CT image pairs have not been established yet. In this paper, we propose a dental CT image denoising method based on the transfer learning of a generative adversarial network (GAN) from the public-domain CT images. We trained a generative adversarial network with the Wasserstein loss function (WGAN) using 5,100 high- and low-dose medical CT image pairs of human chest and abdomen. For the generative network of GAN, we used the U-net structure of five stages to exploit its high computational efficiency. After training the proposed network, named U-WGAN, we fine-tuned the network with 3,006 dental CT image pairs of two different human skull phantoms. For the high- and low-dose scans of the phantoms, we set the tube current of the dental CT to 10 mA and 4 mA, respectively, with the tube voltage set to 90 kVp in both scans. We applied the trained network to denoising of low-dose dental CT images of dental phantoms and adult humans. The U-net processed images showed over-smoothing effects even though U-net had a good performance in the quantitative metrics. U-WGAN showed similar denoising performance to WGAN, but it reduced the computation time of WGAN by a factor of 10. The fine-tuning procedure in the transfer learning scheme enhanced the network performance in terms of the quantitative metrics, and it also improved visual appearance of the processed images. Even though the number of fine-tuning images was very limited in this study, we think the transfer learning scheme can be a good option for developing deep learning networks for dental CT image denoising.

Design and Implementation of a Desensitization Method of CT Medical Image based on DICOM

Medical big data, medical image processing and other fields are based on a large number of clinical data. Data desensitization is of great significance to the privacy protection of patients. CT image desensitization based on DICOM standard protocol only uses the strategy of hiding or deleting the data on the image, which greatly weakens the value of data use and cannot meet the normal value demand. Data desensitization should consider that it can still meet the needs of data mining and other applications on the basis of ensuring the privacy security of users. Based on the above application requirements, a strategy to anonymize patient identity is developed. Based on this strategy, a desensitization method that can meet the actual desensitization standards and analysis application requirements at the same time is proposed. This method can not only protect the privacy of patients to a great extent, but also ensure the availability and authenticity of desensitized data. In the future, this method has great potential in the application of CT image desensitization.

Medical Image Segmentation Algorithm Based on Optimized Convolutional Neural Network-Adaptive Dropout Depth Calculation

Medical image segmentation is a key technology for image guidance. Therefore, the advantages and disadvantages of image segmentation play an important role in image-guided surgery. Traditional machine learning methods have achieved certain beneficial effects in medical image segmentation, but they have problems such as low classification accuracy and poor robustness. Deep learning theory has good generalizability and feature extraction ability, which provides a new idea for solving medical image segmentation problems. However, deep learning has problems in terms of its application to medical image segmentation: one is that the deep learning network structure cannot be constructed according to medical image characteristics; the other is that the generalizability y of the deep learning model is weak. To address these issues, this paper first adapts a neural network to medical image features by adding cross-layer connections to a traditional convolutional neural network. In addition, an optimized convolutional neural network model is established. The optimized convolutional neural network model can segment medical images using the features of two scales simultaneously. At the same time, to solve the generalizability problem of the deep learning model, an adaptive distribution function is designed according to the position of the hidden layer, and then the activation probability of each layer of neurons is set. This enhances the generalizability of the dropout model, and an adaptive dropout model is proposed. This model better addresses the problem of the weak generalizability of deep learning models. Based on the above ideas, this paper proposes a medical image segmentation algorithm based on an optimized convolutional neural network with adaptive dropout depth calculation. An ultrasonic tomographic image and lumbar CT medical image were separately segmented by the method of this paper. The experimental results show that not only are the segmentation effects of the proposed method improved compared with those of the traditional machine learning and other deep learning methods but also the method has a high adaptive segmentation ability for various medical images. The research work in this paper provides a new perspective for research on medical image segmentation.

An efficient and high quality medical CT image enhancement algorithm

AbstractAiming at the process of medical diagnosis, many problems such as unclear images and low contrast are often caused by noise and interference in the process of medical image acquisition and transmission. This article proposes a new image enhancement method that combines the wavelet domain with the spatial domain. First, we input two identical images (Both of the identical images are original images.) in which the first image is enhanced by histogram equalization. Then, the two images are divided into four sub‐images by a two‐dimensional wavelet transform. The average of the low‐frequency coefficients of the low‐frequency sub‐images of the two images is taken as the low‐frequency coefficients of the final reconstruction. Second, aiming at the problem that the contrast may be too low, the fourth high‐frequency sub‐image is blurred (sharpened) twice. The fourth high‐frequency sub‐image after blurring is denoised by median filtering. Finally, the four sub‐images are fused to obtain the enhanced image. The experimental results show that the peak signal‐to‐noise ratio, structural similarity, and processing time of the proposed algorithm are better than those of other contrast algorithms, especially the processing time. These objective indicators show that the proposed algorithm can not only effectively suppress noise but also significantly enhance the contrast. Subjectively, compared with other algorithms, the proposed algorithm achieves a better visual effect and greatly reduces the processing time.

Whale Optimization for Wavelet-Based Unsupervised Medical Image Segmentation: Application to CT and MR Images

Image segmentation plays crucial role in medical image analysis and forms the basis for clinical diagnosis and patient’s treatment planning. But the large variation in organ shapes, inhomogeneous intensities, poor contrast, organic nature of textures and complex boundaries in medical images makes segmentation process adverse and challenging. Further, the absence of annotated ground-truth dataset in medical field limits the advantages of the trending deep learning techniques causing several setbacks. Though numerous unsupervised methods are reported in literature to combat the challenges in medical domain, achieving better segmentation quality still remains as an open issue. This work aim to address this issue integrating the strength of multiresolution analysis and the meta-heuristic optimization techniques for unsupervised medical image segmentation. The proposed approach employs undecimated wavelet frames to extract translation invariant texture and gray information at different orientations to effectively characterize the textures in medical images. Next, the approach introduces the latest meta-heuristic whale optimization algorithm (WOA), the global optimizer to enhance the performance of unsupervised clustering algorithm with optimized cluster centers to cluster the extracted wavelet texture features. Moreover, the study contributes to fill the gap in literature investigating for the first time different intelligence algorithms such as fuzzy, genetic algorithm (GA) and particle swarm optimization (PSO) for unsupervised medical image segmentation to demonstrate the efficacy of the proposed approach. Evaluation was performed on medical CT and MR images based on feature similarity (FSIM), dice (DC) and feature of merits (FOMs). Experimental results demonstrates the supremacy of the proposed approach over other intelligence algorithms. Finally, statistical study with ANOVA analysis was carried out to confirm the significance of the proposed approach in determining the optimal solution and displaying promising segmentation results toward diagnostic support for radiologist.

Endoscopic Path Planning in Robot-Assisted Endoscopic Nasal Surgery

In robot-assisted endoscopic nasal surgery, due to the slender and complicated nasal cavity and sinus anatomical structure, the nasal endoscope of robot end-effector is prone to injure surrounding tissues during surgical approaches. In order to improve the movement safety of the endoscope in robot-assisted endoscopic nasal surgery, a path planning method for endoscopic surgical approaches based on medical image pixel map searching is proposed in this paper, and safe surgical paths from the nasal entrance point to the operating areas are obtained. Firstly, through the trajectory analysis of nasal endoscope tip during surgery, the path planning requirements of endoscope tip are analyzed and obtained. Then, considering the slender and complicated anatomy of the nasal cavity, a binarized three-dimensional grid map containing the spatial anatomy of the nasal cavity is constructed based on the patient's CT medical image sequence. The endoscopic surgical approaches are searched and optimized by introducing the A-star algorithm, and safe surgical paths from the nasal entrance point to the operating areas are obtained. Finally, a virtual nasal endoscopy system is developed and tested on a head model containing nasal tissue, and the effectiveness of the planned surgical paths is verified by automatic virtual nasal endoscopy browsing experiment.

CT and MRI Medical Image Fusion Using Discrete Wavelet Transform

In these days, using medical image is very important in hospitals. These medical images give a lot of data about human body for example Computed Tomography (CT) identifies the bone structure, Magnetic Resonance Image (MRI) image gives information about tissue data, Positron on Emission Tomography (PET) and Single photon release computed tomography (SPECT) give human body functionality data. but these images can't give clear data image for disease diagnosis and treatment planning. So, these different modality complementary data for effective disease analysis is required. In this work we fused two images (CT and MRI) by using discrete wavelet transform then applied this transform on all types of wavelets (haar, Daubechies, Mexican Hat, Symlets, Morlet, Shannon).

Medical Image Authentication by SWT and SVD

The proposed paper work is implemented using Stationary Wavelet Transformation (SWT) with Singular Value Decomposition (SVD).Even though, there are many other transformations, the Stationary Wavelet Transformation method is chosen for its shift invariance property. The designed method has three steps; the first step is the decomposing of the Medical image into sub-bands using SWT to find the value of sub band and as a second step is to apply SVD, third step will combine both the images with scaling factor. The experiments were conducted over gray scale of MRI and CT Medical images. The statistics of proposed method indicates that imperceptibility of Watermarked Medical images have a Peak Signal to Noise Ratio (PSNR) value of 50 DB for medical images. The robustness is ensured by having Correlation Coefficient (CC) of 1 for the retrieved watermark images. Security for the watermark is extended by encrypting the watermark with chaotic sequence.

Medical low-dose CT image denoising based on variable order variational model

Medical low-dose CT image denoising based on variable order variational model

Quality of Service aware Medical CT Image Transmission Anti-collision Mechanism Based on Big Data Autonomous Anti-collision Control

Background: At present, due to the limitation of hardware, software and network transmission performance, the medical diagnosis of medical CT image equipment is easy to be carried out based on the wrong image. In addition, due to the complex structure of human organs and unpredictable lesion location, it is difficult to judge the reliability of medical CT images, spatial localization of the lesion, two-dimensional slice images and shape based on stereotypes. Therefore, how to improve the efficiency of medical CT terminal and the image quality has become the key technology to improve the satisfaction of medical diagnosis and treatment. Objective: To improve the work efficiency of medical CT terminal and medical image transmission quality, with the medical CT terminal state and service quality. Methods: Firstly, from the view of throughput, packet loss rate, delay and so on, a QoS aware model for medical CT image transmission has been established. Then, with throughput, packet length, path loss, service area size, access point location, and the number of medical CT terminals, the performance change regulation of the medical CT image transmission is completed and the optimal quality of service guarantee parameters sequence is obtained. Next, the medical CT image big data autonomous collision control scheme is proposed. Results: The experimental and mathematical results verify the real-time performance, reliability, effectiveness and feasibility of the proposed medical CT image transmission anti-collision mechanism. Conclusion: The proposed scheme can satisfy the high-quality high demand for data transmission at the same time, according to a variety of user experience demand and real-time adjustment of medical CT terminal working state, which provides effective data quality assurance and optimization of the network source distribution, and also enhances the quality of medical image data transmission service.

Textural Analysis of Liver Tumor using Watershed Segmentation based on Statistical and Geometrical features

The liver diseases can define as the tumor or disorder that can affect the liver and causes deformation in its shape. The early detection and diagnose of the tumor using CT medical images, helps the detector to specify the tumor perfectly. This search aims to detect and classify the liver tumor depending on the use of a computer (image processing and textural analysis) helps in getting an accurate diagnosis. The methods which are used in this search depend on creating a binary mask used to separate the liver from the origins of the other in the CT images. The threshold has been used as an early segmentation. A Process, the watershed process is used as a classification technique to isolate the tumor which is cancer and cyst.
 The test images are taking five for cancer case and five for the cyst, the geometrical and statistical features are calculated for both cases to identify the identity for each case. The statistical features are obtained from the gray level co-occurrence matrix which is contrast, homogeneity, correlation and energy. The geometrical features are the area of the tumor, the circumference and irregularity. The irregularity tends to lose one of the cyst cases than for cancer because of its shape in more regular than cancer.

Medical Image Enhancement to Extract Brain Tumors from CT and MRI images

Always MRI and CT Medical images are noisy so that preprocessing is necessary for enhance these images to assist clinicians and make accurate diagnosis. Firstly, in the proposed method uses two denoising filters (Median and Slantlet) are applied to images in parallel and the best enhanced image gained from both filters is voted by use PSNR and MSE as image quality measurements. Next, extraction of brain tumor from cleaned images is done by segmentation method based on k-mean. The result shows that the proposed method is giving an optimal solution due to denoising method which is based on multiple filter types to obtain best clear images and that is leads to make the extraction of tumor more precision best.

A novel YOLOv3-arch model for identifying cholelithiasis and classifying gallstones on CT images.

Locating diseases precisely from medical images, like ultrasonic and CT images, have been one of the most challenging problems in medical image analysis. In recent years, the vigorous development of deep learning models have greatly improved the accuracy in disease location on medical images. However, there are few artificial intelligent methods for identifying cholelithiasis and classifying gallstones on CT images, since no open source CT images dataset of cholelithiasis and gallstones is available for training the models and verifying their performance. In this paper, we build up the first medical image dataset of cholelithiasis by collecting 223846 CT images with gallstone of 1369 patients. With these CT images, a neural network is trained to “pick up” CT images of high quality as training set, and then a novel Yolo neural network, named Yolov3-arch neural network, is proposed to identify cholelithiasis and classify gallstones on CT images. Identification and classification accuracies are obtained by 10-fold cross-validations. It is obtained that our Yolov3-arch model is with average accuracy 92.7% in identifying granular gallstones and average accuracy 80.3% in identifying muddy gallstones. This achieves 3.5% and 8% improvements in identifying granular and muddy gallstones to general Yolo v3 model, respectively. Also, the average cholelithiasis identifying accuracy is improved to 86.50% from 80.75%. Meanwhile, our method can reduce the misdiagnosis rate of negative samples by the object detection model.

Improving Sparse Compressed Sensing Medical CT Image Reconstruction

For the incomplete scanning data, the traditional algorithms cannot guarantee that the medical CT reconstruction image meets the diagnostic requirements. In this case, a medical CT image reconstruction method with the limited angle projection is proposed. According to the theory of compressed sensing, medical CT images with a sparse representation can be reconstructed from the incomplete scanning data and provide reliable information for the diagnosis. The sparse representation of CT images is performed by the sparse patch-ordering wavelet-tree transform, and the digital features of sparse coefficients are used as regularization terms to ensure the validity of the solution. Meantime, the weighting term is added into the fidelity item to reduce the influence of noise on the reconstruction results. The extended Lagrange method is used to solve the constrained objective function iteratively so as to realize the reconstruction of low dose medical CT images. Simulation results demonstrate that the reconstructed image can not only satisfy the completeness condition of projection data, but also can reconstruct the high quality image and effectively improve the mean square error and the structural similarity index.

CT image reconstruction with SuiteSparseQR factorization package

SuiteSparseQR is a factorization package for sparse matrices oriented to parallelism in multicore architectures. It employs BLAS and LAPACK as well as Intel's Threading Building Blocks to achieve high performance. Through the SPQR method implemented in this package we can use the QR decomposition to reconstruct CT images efficiently. In this paper, we analyze the behavior of the package applied to the reconstruction of medical CT images, studying the quality of the obtained image. To this purpose, we use the image dataset DeepLesion, which provides various CT studies of different lesions in different organs or tissues. We also compare it to our previous iterative reconstruction method called LSQR. This new method is promising since the computations are simplified if we compare it to the iterative options and the reconstructions are high-quality, as the results show.

An Effective Diffeomorphic Model and Its Fast Multigrid Algorithm for Registration of Lung CT Images

Abstract Image registration is the process of aligning sets of similar, but different, intensity image functions to track changes between the images. In medical image problems involving lung images, variational registration models are a very powerful tool which can aid in effective treatment of various lung conditions and diseases. However, a common drawback of many variational models, such as the diffusion model and even optic flow models, is the lack of control of folding in the deformations leading to physically inaccurate transformations. For this reason, such models are generally not suitable for real life lung imaging problems where folding cannot occur. There are two approaches offering reliable solutions (though not necessarily accurate). The first approach is a parametric model such as the affine registration model, still widely used in many applications, but is unable to track local changes or yield accurate results. The second approach is to impose an extra constraint on the transformation of registration at the cost of increased non-linearity. An alternative to the second approach, achieving diffeomorphic transforms without adding any constraints, is an inverse consistent model such as by Christensen and Johnson (2001) from computing explicitly both the forward and inverse transforms. However, one must deal with the strong non-linearity in the formulation. In this paper we first propose a simplified inverse consistent model to avoid the inclusion of strong non-linearities and then a fast non-linear multigrid (NMG) technique to overcome the extra computational work required by the inverse consistent model. Experiments, performed on real medical CT images, show that our proposed inverse consistent model is robust to both parameter choice and non-folding in the transformations when compared with diffusion type models.

Particle swarm optimisation aided weighted averaging fusion strategy for CT and MRI medical images

Multimodal medical image fusion is a technique that combines two or more images into a single output image in order to enhance the accuracy of clinical diagnosis. In this paper, a non-subsampled contourlet transform (NSCT) image fusion framework that combines CT and MRI images is proposed. The proposed method decomposes the source images into low and high frequency bands using NSCT and the information across the bands are combined using weighted averaging fusion rule. The weights are optimised by particle swarm optimisation (PSO) with an objective function that jointly maximises the entropy and minimises root mean square error to give improved image quality, which makes different from existing fusion methods in NSCT domain. The performance of the proposed fusion framework is illustrated using five sets of CT and MRI images and various performance metrics indicate that the proposed method is highly efficient and suitable for medical application in better decision making.

Particle swarm optimisation aided weighted averaging fusion strategy for CT and MRI medical images

Particle swarm optimisation aided weighted averaging fusion strategy for CT and MRI medical images

Local bit plane adjacent neighborhood dissimilarity pattern for medical CT image retrieval

In this article, new feature descriptor local bit plane adjacent neighborhood dissimilarity pattern (LBPANDP) is introduced for CT image retrieval. In the proposed method, the input image is first decomposed into eight binary bit planes. Then in each of the first four most significant bit planes, for each center pixel, the dissimilarity information of its binary neighbors are encoded into a single value. The encoded values of each bit plane is then compared with the intensity of the center pixel and this relationship is finally encoded to devise the LBPANDP descriptor. The proposed descriptor is highly discriminative and considers only the first four most significant bit planes for encoding, which greatly reduces the dimension of the feature vector. Retrieval performance of LBPANDP is investigated on two CT image databases and the results show a significant improvement in retrieval efficiency by LBPANDPover many recent techniques.

Influence of the process parameters on mechanical properties of the final parts obtained by selective laser sintering from PA2200 powder

The paper describes how the process parameters affects the mechanical characteristics of laser selective sintered (SLS) parts used in applications of medical, automotive and aerospace fields. The greatest advantage of the additive manufacturing (AM) technology in the medical field is that it allows the use of the patient's medical CT images to obtain specific implants, providing high benefits for both patients and physicians. Despite its increasing use and advantages, the AM process has a series of problemssuch as: the difficulty in obtaining quality part, process interruption or manufacturing part failure. As such, there have been developed experimental researches in order to establish a correlation between the process parameters and the finished part properties. For this analysis, PA 2200 polyamide specimenswere obtained by SLS and subjected to tensile tests. The results correlate the process parameters, providing proof that the tensile properties of SLS specimen are dependent of orientation, position and preheating temperature. Based on the correlation between the process parameters and properties of the PA2200 polyamide, this paper provides a better understanding of the AM process and allows an anticipation onthe best parameters to be used on different parts, leading the optimizing of component properties for medical applications.