With the development of artificial intelligence technologies, it is possible to use computer to read digital medical images. Because Alzheimer’s disease (AD) has the characteristics of high incidence and high disability, it has attracted the attention of many scholars, and its diagnosis and treatment have gradually become a hot topic. In this paper, a multimodal diagnosis method for AD based on three-dimensional shufflenet (3DShuffleNet) and principal component analysis network (PCANet) is proposed. First, the data on structural magnetic resonance imaging (sMRI) and functional magnetic resonance imaging (fMRI) are preprocessed to remove the influence resulting from the differences in image size and shape of different individuals, head movement, noise, and so on. Then, the original two-dimensional (2D) ShuffleNet is developed three-dimensional (3D), which is more suitable for 3D sMRI data to extract the features. In addition, the PCANet network is applied to the brain function connection analysis, and the features on fMRI data are obtained. Next, kernel canonical correlation analysis (KCCA) is used to fuse the features coming from sMRI and fMRI, respectively. Finally, a good classification effect is obtained through the support vector machines (SVM) method classifier, which proves the feasibility and effectiveness of the proposed method.

/pdf/assisted-diagnosis-of-alzheimer-s-disease-based-on-deep-5am1885gi6.pdf

Assisted Diagnosis of Alzheimer’s Disease Based on Deep Learning and Multimodal Feature Fusion

Real-time application based CNN architecture for automatic USCT bone image segmentation

The ordinary people cannot have the capability to detect and classify the brain tumor,9 so the radiologists or the clinical experts are the only person who can detect the encephalon tumor due to their practice and awareness. If the manual detection process will be carried out by the clinical experts, then they will face a lot of problem like time delay, classification accuracy etc., so for this reason, automatic encephalon tumor detection and classification process are required. In this paper, we have detected the encephalon tumor from three schemes, that is, initially, we have collected some raw data and put into the machine learning model; secondly, the classification process is done by extreme learning machine (ELM); and finally, the tumors are extracted from most commonly method gray level co-occurrence method (GLCM). For detecting the tumor region, we have used the one and only common method magnetic resonance image which have also some noises or the doted pictures are available which give a clear image to the clinical experts for detection process. In this work, the classification accuracies have produced as an approximation of 98.09% which has produced the better outcome as compared to the studies that are presented in the related work. From this work, the result is shown as an effective manner for the radiologists as it uses the deep learning method. 

Detection and Classification of Encephalon Tumor Using Extreme Learning Machine Learning Algorithm Based on Deep Learning Method

The advanced world has been influenced by a great deal of sicknesses, and step by step new ones continues arising. Yet, no infection has endangered the humankind as much as ovarian cancer disease. It has taken more women living souls than some other illness. Ovarian cancers principally taking place from the ovaries, generally, in the female’s reproductive glands of ovaries have been encountered tumour. The most important aim of the proposed work is ultrasound ovarian tumour segmentation and subsequently apply ultrasound image is in step by step approach, namely pre-processing and segmentation. In the pre-processing stage, noise such as Speckle Noise, Adaptive Anisotropic Diffusion Filter and Contrast Limited Adaptive Histogram Equalization (CLAHE) used to improve Contrast Adjustment of Ultrasound tumour images and in the segmentation stage, Ovarian tumour are segmented using the GrabCut and FL-SNN Model algorithm. After that, to extract features from Ultrasound tumour image, which is used by statistical texture feature of Gray level Co-occurrence Matrix (GLCM). At long last, Hybrid of GrabCut method and Fuzzy Logic -SNN Model algorithm are processed extracted features of ovarian tumour and this method efficiently segments the tumour area. In the proposed model the subsequent parameters such as bias and weight play a significant part in segmenting the area of tumour. Consequently, it is optimizing the weight parameter value using a Whale search optimisation algorithm. This proposed system gives better and efficient result over existing method.

Performance Analysis of Ultrasound Ovarian Tumour Segmentation Using GrabCut and FL-SNNM

Artificial intelligence (AI) is the capability of training machines to perform detection, cognitive, and diagnostic tasks on a par with humans. AI has evolved to the point that it can perform some medical diagnosis tasks similarly to human experts. We have reviewed the latest AI research in the immense field of medical science, covering various architectures and approaches, with special attention given to brain tumor analysis. This chapter discusses various deep learning architectures used to diagnose brain tumors and compares results with existing architectures. The scope of AI in medical diagnosis tasks, observations and conclusions, and future directions are also discussed. Case studies in medical diagnostics are examined, from basic clustering techniques such as K-means clustering to fuzzy, neurotrophic C-means clustering techniques and kernel graph cuts (KGCs), to advanced artificial intelligence techniques such as deep convolution neural networks (DCNs), atrous convolution neural networks (ACNs), and U-Net architectures, in order to find the area of interest in the coherent/incoherent regions. In the process, we have studied the practical advantages of using atrous convolution and its transpose in the U-Net architecture with attention mechanism instead of traditional convolution and other machine-learning approaches.

A review of deep learning models for medical diagnosis

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.

/pdf/medical-image-segmentation-algorithm-based-on-optimized-54ds5gzmwh.pdf

Jun-e Liu

Papers

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