Remarkable results have been realized by the U-Net network in the task of medical image segmentation. In recent years, many scholars have been researching the network and expanding its structure, such as improvement of encoder and decoder and improvement of skip connection. Based on the optimization of U-Net structure and its medical image segmentation techniques, this paper elucidates in the following: First, the paper elaborates on the application of U-Net in the field of medical image segmentation; Then, the paper summarizes the seven improvement mechanism of U-Net: dense connection mechanism, residual connection mechanism, multi-scale mechanism, ensemble mechanism, dilated mechanism, attention mechanism, and transformer mechanism; Finally, the paper states the ideas and methods on the U-Net structure improvement in a bid to provide a reference for later researches, which plays a significant part in advancing U-Net.
Image Processing, Computer-Assisted/methods*

OBJECTIVEDue to random sampling of non-adaptive, high-quality reconstruction of the original signal, one-dimensional non-stationary multi-channel EEG signal can be achieved automatic detection and analysis.

METHODSA new multicomponent redundant dictionaries with the atoms of the Gaussian function and its first and second derivatives was built in the paper, and reconstructed signal base on compressed sensing measurement model.

RESULTSThe selected dictionary atoms can more effectively match the EEG signals in a variety of transient characteristics of the waveform, allowing the formation of EEG signal is more sparse matching pursuit decomposition. With the theory based on compressed sensing signal sampling, only half of the original signal with different sample size can be used to reconstruct the original signal quality, the important instantaneous features of the waveform can well be maintained.

CONCLUSIONSignal sampling based on the theory of compressed sensing contains enough information of the original signal, using the prior conditions of EEG signals (or compressibility) sparsity, high-dimensional signal and original image can be reconstructed through a certain decoding of linear or nonlinear model.

Machine Learning ; Image Processing, Computer-Assisted/methods*

Image analysis plays an important role in morphological study of tissues and cells and their quantitative analysis. It also contributes to clinical pathological diagnosis. With the rapid development of computer technology great progress has been made in the image analysis system, its measurement, rapidity, accuracy and the extent of automation have been greatly enhanced. In the meantime advances in medicine give impetus to its improvement. In this paper, the process of development, the basic structure and mechanism of image analysis system are discussed and the application of image analysis system to the study of biomedicinal is presented.
Image Processing, Computer-Assisted ; instrumentation ; methods ; Medicine

After a study on the characteristic of ECG data, we propose here in this paper a lossless compression method of ECG data, which is based on JPEG2000. It integrates both 1D and 2D compression. The method has been verified through all forty-eight records in MIT-BIH Arrhythmia database. And the result shows that the method has a better compression rate and a good computational efficiency.
Algorithms ; Electrocardiography ; methods ; Image Processing, Computer-Assisted

The high frame rate (HFR) ultrasonic imaging system based on linear frequency-modulated (LFM) signal constructs images at a high frame rate; the signal-to-noise ratio (SNR) of this system can also be improved. Unfortunately, such pulse compression methods that increase the SNR usually cause range sidelobe artifacts. In an imaging situation, the effects of the sidelobes extending on either side of the compressed pulse will be self-noise along the axial direction and masking of weaker echoes. The improvement on high frame rate ultrasonic imaging system based on LFM signal is considered in this paper. In this proposed scheme, a predistorted LFM signal is used as excited signal and a mismatched filter is applied on receiving end. The results show that the proposed HFR ultrasonic imaging system can achieve higher SNR and the axial resolution is also improved.
Algorithms ; Humans ; Image Enhancement ; methods ; Image Processing, Computer-Assisted ; methods ; Signal Processing, Computer-Assisted ; Ultrasonography ; methods

The scale of local structure is a key parameter in medical image registration. Unfortunately, no much attention has been paid to the scale selection for the local structures in the images. This paper proposes a data-driven scale selection method for local structures in the image. By using minimal description length criterion to maximize the posterior probability of local structure region with coherence constraint based on the Markov random field model, an optimal scale for each local structure, which is segmented with super-pixel representation, is assigned in terms of variance in a discrete anisotropic scale space. Therefore, the local structure's scale can be selected for further non-rigid medical image registration.
Algorithms ; Image Enhancement ; methods ; Image Processing, Computer-Assisted ; methods

Lower contrast and speckle noise are the main reasons which decline the quality of medical ultrasonic images. In this paper, a new method is proposed to filter the speckle noises and enhance the contrast simultaneously. Anisotropic diffusion filtering method was firstly applied to filter images. Then the loss of information, which the contrast function of contrast enhancement model lies on, was obtained. Finally, the contrast can be enhanced by using enhancement model. Experimental results show that the proposed method not only removes the speckle noises effectively, but also enhances the contrast obviously. This method supplies an effective approach for improving the quality of medical ultrasonic images.
Algorithms ; Anisotropy ; Humans ; Image Enhancement ; Image Processing, Computer-Assisted ; methods ; Signal Processing, Computer-Assisted ; Ultrasonography ; methods

In deep learning-based image registration, the deformable region with complex anatomical structures is an important factor affecting the accuracy of network registration. However, it is difficult for existing methods to pay attention to complex anatomical regions of images. At the same time, the receptive field of the convolutional neural network is limited by the size of its convolution kernel, and it is difficult to learn the relationship between the voxels with far spatial location, making it difficult to deal with the large region deformation problem. Aiming at the above two problems, this paper proposes a cascaded multi-level registration network model based on transformer, and equipped it with a difficult deformable region perceptron based on mean square error. The difficult deformation perceptron uses sliding window and floating window techniques to retrieve the registered images, obtain the difficult deformation coefficient of each voxel, and identify the regions with the worst registration effect. In this study, the cascaded multi-level registration network model adopts the difficult deformation perceptron for hierarchical connection, and the self-attention mechanism is used to extract global features in the basic registration network to optimize the registration results of different scales. The experimental results show that the method proposed in this paper can perform progressive registration of complex deformation regions, thereby optimizing the registration results of brain medical images, which has a good auxiliary effect on the clinical diagnosis of doctors.
Algorithms ; Neural Networks, Computer ; Image Processing, Computer-Assisted/methods*

Visual prostheses have the potential to restore partial vision for the blinds. The stimulating electrodes generate reproducible phosphenes. Still limited by the low resolution vision used in visual prostheses nowadays, it is important to optimize the image processing strategies in order to deliver better visual information to the patients. This paper presents a review of the current research progress on the image processing strategies based on visual attention models under simulated prosthetic vision and related psychophysics.
Computer Simulation ; Image Processing, Computer-Assisted ; methods ; Visual Prosthesis