Narrow Results

In medical science, appropriate monitoring of the function of the human body, physical representations, and accurate disease is a complex task. It helps in diagnosing and treating diseases by revealing internal structures hidden by the skin and bones. Image enhancement techniques play a crucial role in medical imaging, as they improve diagnostic accuracy and visualization. However, the conventional methods were limited by noise sensitivity, lack of adaptability, and computational complexity. Therefore, this paper proposes One-to-One Honey Optimization (OOHO) for medical image enhancement. At first, the input brain Magnetic Resonance Imaging (MRI) image is processed to find the histogram. After that, histogram clipping is performed based on the clipping threshold. Then, the exposure threshold is computed to create sub-histograms. Moreover, the Probability Density Function (PDF) is computed and updated. Likewise, the Cumulative Density Function (CDF) and the mapping function are computed in each sub-histogram. Therefore, the finest enhanced image is obtained by integrating all sub-images. The fitness value is computed utilizing the cost function. Furthermore, the optimal threshold is accomplished by utilizing OOHO, which is developed by the integration of a One-to-One-based Optimizer (OOBO) and Honey Badger optimization (HBO). The evaluation results show that the OOHO attained a Degree of Distortion (DD) of 0.270, a Pear Signal to Noise Ratio (PSNR) of 47.060 dB, a Mean Square Error (MSE) of 0.570, Structural Similarity Index Measure (SSIM) as 0.960, and Root Mean Square Error (RMSE) of 0.755. The high performance is noted by the devised model and the MSE of the OOBO, HBO, hybrid Simulated Annealing-Evaporation Rate-based Water Cycle (SA-ERWCA) algorithm, Particle Swarm Optimization combined with Histogram Equalization (PSO-HE), Genetic Algorithm-based Adaptive Histogram Equalization (GAAHE), High-Quality Guidance Network (HQG-Net), multi-scale attention generative adversarial network (MAGAN), and Gaussian Quantum-behaved Arithmetic Optimization Algorithm (GQAOA) methods are 0.657, 0.637, 0.637, 0.627, 0.670, 0.650, 0.650, and 0.640, respectively.

This paper proposes a high-capacity reversible watermarking algorithm for medical image analysis based on the difference block histogram, which will benefit the medical image authentication and doctor–patient confidentiality. By dividing the original medical image into blocks, the method displaces the peak point of a block's histogram of difference and embeds multi-bit information at 1 pixel point. In so doing, secret communication and storage of large-capacity invisible medical diagnoses and patients personal confidential data can be achieved. Once the watermark is extracted, not only the image integrity is authenticated, but also the original image and personal data of the patient can be recovered in a nondestructive way. With low computational complexity, a high embedding capacity and little demand for auxiliary information, the proposed algorithm is highly secure and practical.

Due to the rapid rise of telemedicine, a lot of patients’ information will be transmitted through the Internet. However, the patients’ information is related to personal privacy, therefore, patients’ information needs to be encrypted when transmited and stored. Medical image encryption is a part of it. Due to the informative fine features of medical images, a common image encryption algorithm is no longer applied. Common encryption algorithm has a single theory based on chaos image encryption algorithm, other encryption algorithms are based on information entropy. However, the images processed with these cipher text encryption algorithm are cyclical, the outline is clear and the anti-tamper capability is not strong. In view of the bit being the smallest measure unit of pixel, in order to overcome the weakness from above algorithm, and take the advantage of the chaotic system, this paper will present a chaotic medical image encryption algorithm based on bit-plane decomposition. The paper combines the image encryption and chaotic system to improve the security. This way, it can increase the security of key space and image effectively. The histogram, pixel correlation, number of pixels change rate (NPCR) and other experimental results show that the algorithm satisfies the desired effect.

Taking the improved ant colony algorithm based on bacterial chemotaxis as a means, this paper proposes one new swarm intelligence optimization algorithm to solve the medical image edge detection problem. The improved ant colony algorithm based on bacterial chemotaxis mainly aims at the shortcoming that the basic ant colony algorithm lacks initial pheromone, and combines bacterial chemotaxis algorithm with basic ant colony algorithm. Firstly, feasible better solution can be found through bacterial chemotaxis algorithm and fed back as initial pheromone. Then ant colony algorithm is implemented to search for the global optimal solution. The algorithm test indicates that the improved ant colony algorithm is more effective in the aspects of searching precision, reliability, optimization speed and stability compared with basic ant colony algorithm. Finally, the improved ant colony algorithm is applied into the edge detection of medical image. It can be seen from the computer simulation that compared with other operators and basic ant colony algorithm on the issue of solving medical image edge detection, the improved ant colony algorithm has superiority and the detected edge is clearer.

The increasing diverse demand for image feature recognition and complicated relationships among image pixels cannot be fully and effectively handled by traditional single image recognition methods. In order to effectively improve classification accuracy in image processing, a deep belief network (DBN) classification model based on probability measure rough set theory is proposed in our research.

First, the incomplete and inaccurate fuzzy information in the original image is preprocessed by the rough set method based on probability measure. Second, the attribute features of the image information are extracted, the attribute feature set is reduced to generate the classification rules, and key components are extracted as the input of the DBN. Third, the network structure of the DBN is determined by the extracted classification rules, and the importance of the rough set attributes is integrated and the weights of the neuronal nodes are corrected by the backpropagation (BP) algorithm. Last, the DBN is trained to classify images. The experimental analysis of the proposed method for medical imagery shows that it is more effective than current single rough set approach or the taxonomy of deep learning.

The traditional watershed algorithm has the limitation of false mark in medical image segmentation, which causes over-segmentation and images to be contaminated by noise possibly during acquisition. In this study, we proposed an improved watershed segmentation algorithm based on morphological processing and total variation model (TV) for medical image segmentation. First of all, morphological gradient preprocessing is performed on MRI images of brain lesions. Secondly, the gradient images are denoised by the all-variational model. While retaining the edge information of MRI images of brain lesions, the image noise is reduced. And then, the internal and external markers are obtained by forced minimum technique, and the gradient amplitude images are corrected by using these markers. Finally, the modified gradient image is subjected to watershed transformation. The experiment of segmentation and simulation of brain lesion MRI image is carried out on MATLAB. And the segmentation results are compared with other watershed algrothims. The experimental results demonstrate that our method obtains the least number of regions, which can extract MRI images of brain lesions effectively. In addition, this method can inhibit over-segmentation, improving the segmentation results of lesions in MRI images of brain lesions.

Because of the limitations of convolution kernel, the traditional image segmentation network is not sufficient to obtain the context information, but the image segmentation task is very dependent on the context information. Transformer’s linear input can just get enough context information. In this paper, we propose a transformer segmentation network hyperfusion transformer based on a pyramid structure. First, the model divides the single-scale coding form into several-different-scale coding forms, and then fuses the decoding results. Second, in order to ensure the specificity of the output characteristics of each branch, we orthogonalize the results of a variety of different scales. By orthogonalizing in pairs, we can ensure that the results obtained by different branches are not completely similar to a certain extent, and reduce the redundancy of branch information. On the two datasets, the method in this paper surpasses a variety of classical models under multiple evaluation indexes, confirming that it is an effective segmentation method.

In the field of medical diagnostics, interested parties have resorted increasingly to medical imaging. It is well established that the accuracy and completeness of diagnosis are initially connected with the image quality, but the quality of the image is itself dependent on a number of factors including primarily the processing that an image must undergo to enhance its quality. The quality evaluation of compressed image is necessary to judge the performance of a compression method. This paper introduces an algorithm for medical image compression based on hybrid nonsubsampled contourlet (NSCT) and quincunx wavelet transforms (QWT) coupled with set partitioning in hierarchical trees (SPIHT) coding algorithm, of which we present the objective measurements (PSNR, EDGE, WPSNR, MSSIM, VIF, and WSNR) in order to evaluate the quality of the image.

In the last years, the internet and multimedia technologies are widely used to exchange medical images and practicing telemedicine; which exposes this data to various illegitimate attacks due to their sensitivity. For this reason, many researchers work to afford a proper security and an efficient protection during the transmission. In this context, we propose a blind multi-watermarking approach for medical images using Lifting Wavelet Transform and Fast Walsh–Hadamard Transform. This approach proves the ability to embed a binary image ($32\times 32$), used for integrity purposes, in addition to the patient information (EPR) associated with the medical cover image. Moreover, an error correcting code was used to improve the security of the EPR. We tested the proposed approach on different imaging modalities, and the obtained results show a good imperceptibility and robustness against several common attacks.

Medical image segmentation plays a crucial role in clinical diagnosis and therapy systems, yet still faces many challenges. Building on convolutional neural networks (CNNs), medical image segmentation has achieved tremendous progress. However, owing to the locality of convolution operations, CNNs have the inherent limitation in learning global context. To address the limitation in building global context relationship from CNNs, we propose LGNet, a semantic segmentation network aiming to learn local and global features for fast and accurate medical image segmentation in this paper. Specifically, we employ a two-branch architecture consisting of convolution layers in one branch to learn local features and transformer layers in the other branch to learn global features. LGNet has two key insights: (1) We bridge two-branch to learn local and global features in an interactive way; (2) we present a novel multi-feature fusion model (MSFFM) to leverage the global contexture information from transformer and the local representational features from convolutions. Our method achieves state-of-the-art trade-off in terms of accuracy and efficiency on several medical image segmentation benchmarks including Synapse, ACDC and MOST. Specifically, LGNet achieves the state-of-the-art performance with Dice’s indexes of 80.15% on Synapse, of 91.70% on ACDC, and of 95.56% on MOST. Meanwhile, the inference speed attains at 172 frames per second with $224\times 224$ input resolution. The extensive experiments demonstrate the effectiveness of the proposed LGNet for fast and accurate for medical image segmentation.

Breast cancer (BC) is one of the major principal sources of high mortality among women worldwide. Consequently, early detection is essential to save lives. BC can be diagnosed with different modes of medical images such as mammography, ultrasound, computerized tomography, biopsy, and magnetic resonance imaging. A histopathology study (biopsy) that results in images is often performed to help diagnose and analyze BC. Transfer learning (TL) is a machine-learning (ML) technique that reuses a learning method that is initially built for a task to be applied to a model for a new task. TL aims to enhance the assessment of desired learners by moving the knowledge contained in another but similar source domain. Consequently, the challenge of the small dataset in the desired domain is reduced to build the desired learners. TL plays a major role in medical image analysis because of this immense property. This paper focuses on the use of TL methods for the investigation of BC image classification and detection, preprocessing, pretrained models, and ML models. Through empirical experiments, the EfficientNets pretrained neural network architectures and ML classification models were built. The support vector machine and eXtreme Gradient Boosting (XGBoost) were learned on the BC dataset. The result showed a comparative but good performance of EfficientNetB4 and XGBoost. An outcome based on accuracy, recall, precision, and F1_Score for XGBoost is 84%, 0.80, 0.83, and 0.81, respectively.

In this paper, the tilt correction of computerized tomography (CT) and magnetic resonance (MR) medical images is the main point of interest. The centroid of the medical subimage is calculated, and the rotation angle α is obtained by separately using two methods: singular value decomposition (SVD) and principal component analysis (PCA), respectively. In addition, the whole medical image is rotated around the centroid by -α to correct the tilt. Based on this, according to the uniformity of the medical subimage the rotation angle α is further adjusted, which achieves better correction effect and performance. The experimental results show that the correction effect of SVD is the same as that of PCA, the proposed methods are fairly reliable and accurate for the determination of tilt angles, and are practical and effective tilt correction techniques. In addition, the correction methods are regarded as the preprocesses of image registration, and hence are used to get the registering parameters by incorporating the pattern search method. The experimental results reveal that they can significantly reduce the computational load, accurately get transformation parameters, and overcome the problem of easily getting into the local optimum.

In this paper, the $\alpha$-Renyi entropy and $\alpha$-Renyi-based mutual information (RMI) are first introduced. Then the influence of the parameter $\alpha$ on the curve of the RMI and the computational load of image registration are discussed and analyzed to explore the appropriate parameter ranges. Finally, the RMI with the appropriate parameter $\alpha$ is viewed as the similarity measure between the reference and floating images. In the experiments, the Simplex method is chosen as the multi-parameter optimization one. The experimental results reveal that the proposed method has low computational load, fast registration and good registration accuracy. It is adapted to both mono-modality and multi-modality image registrations.

In this paper, a new method for medical image edge detection which is based on the first derivative and zero crossings of second derivative knowledge is proposed. Although there are many traditional methods for image edge detection, especially some methods based on zero crossings, most of them are sensitive to noise so that the image edge can’t be detected clearly. Here this new algorithm is given to avoid such disadvantages. Our method computes the second derivative to get zero crossings, and calculates the first derivative to find out the local maxima, and then deletes those whose second derivative is zero crossings but first derivative is not local maxima which means that maybe these edges are produced by noise. The experiment results show that this new method not only has perfectible result of edge detection but also has good robustness to noise.

In this paper, a method for the boundary detection of breast tumors from ultrasound (US) is proposed. In the preprocessing, the total-variation filter is used to reduce the speckles. The method is based on the active contour model using a region based approach and parametric representation of the contour. Determination of the real tumor boundary with the active contour model requires an initial contour estimate. We apply an automatic initial contour-finding method based on automatic thresholding that not only maintains the tumor shape, but also is close to the tumor boundary. For measuring the accuracy of the model, three different error measures were used in comparison with the delineations of an expert.