Exploiting Deep Learning Techniques for Colon Polyp Segmentation

As colon cancer is among the top causes of death, there is a growing interest in developing improved techniques for the early detection of colon polyps. Given the close relation between colon polyps and colon cancer, their detection helps avoid cancer cases. The increment in the availability of colorectal screening tests and the number of colonoscopies have increased the burden on the medical personnel. In this article, the application of deep learning techniques for the detection and segmentation of colon polyps in colonoscopies is presented. Four techniques were implemented and evaluated: Mask-RCNN, PANet, Cascade R-CNN and Hybrid Task Cascade (HTC). These were trained and tested using CVC-Colon database, ETIS-LARIB Polyp, and a proprietary dataset. Three experiments were conducted to assess the techniques performance: 1) Training and testing using each database independently, 2) Mergingd the databases and testing on each database independently using a merged test set, and 3) Training on each dataset and testing on the merged test set. In our experiments, PANet architecture has the best performance in Polyp detection, and HTC was the most accurate to segment them. This approach allows us to employ Deep Learning techniques to assist healthcare professionals in the medical diagnosis for colon cancer. It is anticipated that this approach can be part of a framework for a semi-automated polyp detection in colonoscopies.     

Joint Deep Matching Model of OCT Retinal Layer Segmentation

Optical Coherence Tomography (OCT) is very important in medicine and provide useful diagnostic information. Measuring retinal layer thicknesses plays a vital role in pathophysiologic factors of many ocular conditions. Among the existing retinal layer segmentation approaches, learning or deep learning-based methods belong to the state-of-art. However, most of these techniques rely on manual-marked layers and the performances are limited due to the image quality. In order to overcome this limitation, we build a framework based on gray value curve matching, which uses depth learning to match the curve for semi-automatic segmentation of retinal layers from OCT. The depth convolution network learns the column correspondence in the OCT image unsupervised. The whole OCT image participates in the depth convolution neural network operation, compares the gray value of each column, and matches the gray value sequence of the transformation column and the next column. Using this algorithm, when a boundary point is manually specified, we can accurately segment the boundary between retinal layers. Our experimental results obtained from a 54-subjects database of both normal healthy eyes and affected eyes demonstrate the superior performances of our approach.     

Skin Lesion Segmentation and Classification Using Conventional and Deep Learning Based Framework

Background: In medical image analysis, the diagnosis of skin lesions remains a challenging task. Skin lesion is a common type of skin cancer that exists worldwide. Dermoscopy is one of the latest technologies used for the diagnosis of skin cancer. Challenges: Many computerized methods have been introduced in the literature to classify skin cancers. However, challenges remain such as imbalanced datasets, low contrast lesions, and the extraction of irrelevant or redundant features. Proposed Work: In this study, a new technique is proposed based on the conventional and deep learning framework. The proposed framework consists of two major tasks: lesion segmentation and classification. In the lesion segmentation task, contrast is initially improved by the fusion of two filtering techniques and then performed a color transformation to color lesion area color discrimination. Subsequently, the best channel is selected and the lesion map is computed, which is further converted into a binary form using a thresholding function. In the lesion classification task, two pre-trained CNN models were modified and trained using transfer learning. Deep features were extracted from both models and fused using canonical correlation analysis. During the fusion process, a few redundant features were also added, lowering classification accuracy. A new technique called maximum entropy score-based selection (MESbS) is proposed as a solution to this issue. The features selected through this approach are fed into a cubic support vector machine (C-SVM) for the final classification. Results: The experimental process was conducted on two datasets: ISIC 2017 and HAM10000. The ISIC 2017 dataset was used for the lesion segmentation task, whereas the HAM10000 dataset was used for the classification task. The achieved accuracy for both datasets was 95.6% and 96.7%, respectively, which was higher than the existing techniques.     

Deep Learning in DXA Image Segmentation

Many existing techniques to acquire dual-energy X-ray absorptiometry (DXA) images are unable to accurately distinguish between bone and soft tissue. For the most part, this failure stems from bone shape variability, noise and low contrast in DXA images, inconsistent X-ray beam penetration producing shadowing effects, and person-to-person variations. This work explores the feasibility of using state-of-the-art deep learning semantic segmentation models, fully convolutional networks (FCNs), SegNet, and U-Net to distinguish femur bone from soft tissue. We investigated the performance of deep learning algorithms with reference to some of our previously applied conventional image segmentation techniques (i.e., a decision-tree-based method using a pixel label decision tree [PLDT] and another method using Otsu’s thresholding) for femur DXA images, and we measured accuracy based on the average Jaccard index, sensitivity, and specificity. Deep learning models using SegNet, U-Net, and an FCN achieved average segmentation accuracies of 95.8%, 95.1%, and 97.6%, respectively, compared to PLDT (91.4%) and Otsu’s thresholding (72.6%). Thus we conclude that an FCN outperforms other deep learning and conventional techniques when segmenting femur bone from soft tissue in DXA images. Accurate femur segmentation improves bone mineral density computation, which in turn enhances the diagnosing of osteoporosis.     

Tissue Segmentation in Nasopharyngeal CT Images Using TwoStage Learning

Tissue segmentation is a fundamental and important task in nasopharyngeal images analysis. However, it is a challenging task to accurately and quickly segment various tissues in the nasopharynx region due to the small difference in gray value between tissues in the nasopharyngeal image and the complexity of the tissue structure. In this paper, we propose a novel tissue segmentation approach based on a two-stage learning framework and U-Net. In the proposed methodology, the network consists of two segmentation modules. The first module performs rough segmentation and the second module performs accurate segmentation. Considering the training time and the limitation of computing resources, the structure of the second module is simpler and the number of network layers is less. In addition, our segmentation module is based on U-Net and incorporates a skip structure, which can make full use of the original features of the data and avoid feature loss. We evaluated our proposed method on the nasopharyngeal dataset provided by West China Hospital of Sichuan University. The experimental results show that the proposed method is superior to many standard segmentation structures and the recently proposed nasopharyngeal tissue segmentation method, and can be easily generalized across different tissue types in various organs.     

Survey on Segmentation and Classification Techniques of Satellite Images by Deep Learning Algorithm

This survey paper aims to show methods to analyze and classify field satellite images using deep learning and machine learning algorithms. Users of deep learning-based Convolutional Neural Network (CNN) technology to harvest fields from satellite images or generate zones of interest were among the planned application scenarios (ROI). Using machine learning, the satellite image is placed on the input image, segmented, and then tagged. In contemporary categorization, field size ratio, Local Binary Pattern (LBP) histograms, and color data are taken into account. Field satellite image localization has several practical applications, including pest management, scene analysis, and field tracking. The relationship between satellite images in a specific area, or contextual information, is essential to comprehending the field in its whole.     

Automated Deep Learning Empowered Breast Cancer Diagnosis Using Biomedical Mammogram Images

Biomedical image processing is a hot research topic which helps to majorly assist the disease diagnostic process. At the same time, breast cancer becomes the deadliest disease among women and can be detected by the use of different imaging techniques. Digital mammograms can be used for the earlier identification and diagnostic of breast cancer to minimize the death rate. But the proper identification of breast cancer has mainly relied on the mammography findings and results to increased false positives. For resolving the issues of false positives of breast cancer diagnosis, this paper presents an automated deep learning based breast cancer diagnosis (ADL-BCD) model using digital mammograms. The goal of the ADL-BCD technique is to properly detect the existence of breast lesions using digital mammograms. The proposed model involves Gaussian filter based pre-processing and Tsallis entropy based image segmentation. In addition, Deep Convolutional Neural Network based Residual Network (ResNet 34) is applied for feature extraction purposes. Specifically, a hyper parameter tuning process using chimp optimization algorithm (COA) is applied to tune the parameters involved in ResNet 34 model. The wavelet neural network (WNN) is used for the classification of digital mammograms for the detection of breast cancer. The ADL-BCD method is evaluated using a benchmark dataset and the results are analyzed under several performance measures. The simulation outcome indicated that the ADL-BCD model outperforms the state of art methods in terms of different measures.     

Investigation on the Chinese Text Sentiment Analysis Based on Convolutional Neural Networks in Deep Learning

Nowadays, the amount of wed data is increasing at a rapid speed, which presents a serious challenge to the web monitoring. Text sentiment analysis, an important research topic in the area of natural language processing, is a crucial task in the web monitoring area. The accuracy of traditional text sentiment analysis methods might be degraded in dealing with mass data. Deep learning is a hot research topic of the artificial intelligence in the recent years. By now, several research groups have studied the sentiment analysis of English texts using deep learning methods. In contrary, relatively few works have so far considered the Chinese text sentiment analysis toward this direction. In this paper, a method for analyzing the Chinese text sentiment is proposed based on the convolutional neural network (CNN) in deep learning in order to improve the analysis accuracy. The feature values of the CNN after the training process are nonuniformly distributed. In order to overcome this problem, a method for normalizing the feature values is proposed. Moreover, the dimensions of the text features are optimized through simulations. Finally, a method for updating the learning rate in the training process of the CNN is presented in order to achieve better performances. Experiment results on the typical datasets indicate that the accuracy of the proposed method can be improved compared with that of the traditional supervised machine learning methods, e.g., the support vector machine method.     

Effectiveness of Deep Learning Models for Brain Tumor Classification and Segmentation

A brain tumor is a mass or growth of abnormal cells in the brain. In children and adults, brain tumor is considered one of the leading causes of death. There are several types of brain tumors, including benign (non-cancerous) and malignant (cancerous) tumors. Diagnosing brain tumors as early as possible is essential, as this can improve the chances of successful treatment and survival. Considering this problem, we bring forth a hybrid intelligent deep learning technique that uses several pre-trained models (Resnet50, Vgg16, Vgg19, U-Net) and their integration for computer-aided detection and localization systems in brain tumors. These pre-trained and integrated deep learning models have been used on the publicly available dataset from The Cancer Genome Atlas. The dataset consists of 120 patients. The pre-trained models have been used to classify tumor or no tumor images, while integrated models are applied to segment the tumor region correctly. We have evaluated their performance in terms of loss, accuracy, intersection over union, Jaccard distance, dice coefficient, and dice coefficient loss. From pre-trained models, the U-Net model achieves higher performance than other models by obtaining 95% accuracy. In contrast, U-Net with ResNet-50 outperforms all other models from integrated pre-trained models and correctly classified and segmented the tumor region.     

Multimodal Fused Deep Learning Networks for Domain Specific Image Similarity Search

The exponential increase in data over the past few years, particularly in images, has led to more complex content since visual representation became the new norm. E-commerce and similar platforms maintain large image catalogues of their products. In image databases, searching and retrieving similar images is still a challenge, even though several image retrieval techniques have been proposed over the decade. Most of these techniques work well when querying general image databases. However, they often fail in domain-specific image databases, especially for datasets with low intraclass variance. This paper proposes a domain-specific image similarity search engine based on a fused deep learning network. The network is comprised of an improved object localization module, a classification module to narrow down search options and finally a feature extraction and similarity calculation module. The network features both an offline stage for indexing the dataset and an online stage for querying. The dataset used to evaluate the performance of the proposed network is a custom domain-specific dataset related to cosmetics packaging gathered from various online platforms. The proposed method addresses the intraclass variance problem with more precise object localization and the introduction of top result reranking based on object contours. Finally, quantitative and qualitative experiment results are presented, showing improved image similarity search performance.     

Ore Image Segmentation Method Based on U-Net and Watershed

Ore image segmentation is a key step in an ore grain size analysis based on image processing. The traditional segmentation methods do not deal with ore textures and shadows in ore images well Those methods often suffer from under-segmentation and over-segmentation. In this article, in order to solve the problem, an ore image segmentation method based on U-Net is proposed. We adjust the structure of U-Net to speed up the processing, and we modify the loss function to enhance the generalization of the model. After the collection of the ore image, we design the annotation standard and train the network with the annotated image. Finally, the marked watershed algorithm is used to segment the adhesion area. The experimental results show that the proposed method has the characteristics of fast speed, strong robustness and high precision. It has great practical value to the actual ore grain statistical task.     

Software Defect Prediction Based on Non-Linear Manifold Learning and Hybrid Deep Learning Techniques

Software defect prediction plays a very important role in software quality assurance, which aims to inspect as many potentially defect-prone software modules as possible. However, the performance of the prediction model is susceptible to high dimensionality of the dataset that contains irrelevant and redundant features. In addition, software metrics for software defect prediction are almost entirely traditional features compared to the deep semantic feature representation from deep learning techniques. To address these two issues, we propose the following two solutions in this paper: (1) We leverage a novel non-linear manifold learning method - SOINN Landmark Isomap (SLIsomap) to extract the representative features by selecting automatically the reasonable number and position of landmarks, which can reveal the complex intrinsic structure hidden behind the defect data. (2) We propose a novel defect prediction model named DLDD based on hybrid deep learning techniques, which leverages denoising autoencoder to learn true input features that are not contaminated by noise, and utilizes deep neural network to learn the abstract deep semantic features. We combine the squared error loss function of denoising autoencoder with the cross entropy loss function of deep neural network to achieve the best prediction performance by adjusting a hyperparameter. We compare the SL-Isomap with seven state-of-the-art feature extraction methods and compare the DLDD model with six baseline models across 20 open source software projects. The experimental results verify that the superiority of SL-Isomap and DLDD on four evaluation indicators.     

Optimized Deep Learning Model for Fire Semantic Segmentation

Recent convolutional neural networks (CNNs) based deep learning has significantly promoted fire detection. Existing fire detection methods can efficiently recognize and locate the fire. However, the accurate flame boundary and shape information is hard to obtain by them, which makes it difficult to conduct automated fire region analysis, prediction, and early warning. To this end, we propose a fire semantic segmentation method based on Global Position Guidance (GPG) and Multi-path explicit Edge information Interaction (MEI). Specifically, to solve the problem of local segmentation errors in low-level feature space, a top-down global position guidance module is used to restrain the offset of low-level features. Besides, an MEI module is proposed to explicitly extract and utilize the edge information to refine the coarse fire segmentation results. We compare the proposed method with existing advanced semantic segmentation and salient object detection methods. Experimental results demonstrate that the proposed method achieves 94.1%, 93.6%, 94.6%, 95.3%, and 95.9% Intersection over Union (IoU) on five test sets respectively which outperforms the suboptimal method by a large margin. In addition, in terms of accuracy, our approach also achieves the best score.     

基于深度强化学习的图像去模糊方法

目的为了有效地去除多种图像模糊,提高图像质量,提出基于深度强化学习的图像去模糊方法。方法选用GoPro与DIV2K这2个数据集进行实验,以峰值信噪比(PSNR)和结构相似性(SSIM)为客观评价指标。通过卷积神经网络获得模糊图像的高维特征,利用深度强化学习结合多种CNN去模糊工具建立去模糊框架,将峰值信噪比(PSNR)作为训练奖励评价函数,来选择最优修复策略,逐步对模糊图像进行修复。结果通过训练与测试,与现有的主流算法相比,文中方法有着更好的主观视觉效果,且PSNR值与SSIM值都有更好的表现。结论实验结果表明,文中方法能有效地解决图像的高斯模糊和运动模糊等问题,并取得了良好的视觉效果,在图像去模糊领域具有一定的参考价值。     

Drug Response Prediction of Liver Cancer Cell Line Using Deep Learning

Cancer is the second deadliest human disease worldwide with high mortality rate. Rehabilitation and treatment of this disease requires precise and automatic assessment of effective drug response and control system. Prediction of treated and untreated cancerous cell line is one of the most challenging problems for precise and targeted drug delivery and response. A novel approach is proposed for prediction of drug treated and untreated cancer cell line automatically by employing modified Deep neural networks. Human hepatocellular carcinoma (HepG2) cells are exposed to anticancer drug functionalized CFO@BTO nanoparticles developed by our lab. Prediction models are developed by modifying ResNet101 and exploiting the transfer learning concept. Last three layers of ResNet101 are re-trained for the identification of drug treated cancer cells. Transfer learning approach in an appropriate choice especially when there is limited amount of annotated data. The proposed technique is validated on acquired 203 fluorescent microscopy images of human HepG2 cells treated with drug functionalized cobalt ferrite@barium titanate (CFO@BTO) magnetoelectric nanoparticles in vitro. The developed approach achieved high prediction with accuracy of 97.5% and sensitivity of 100% and outperformed other approaches. The high performance reveals the effectiveness of the approach. It is scalable and fully automatic prediction approach which can be extended for other similar cell diseases such as lung, brain tumor and breast cancer.     

Intelligent Machine Learning Enabled Retinal Blood Vessel Segmentation and Classification

Automated segmentation of blood vessels in retinal fundus images is essential for medical image analysis. The segmentation of retinal vessels is assumed to be essential to the progress of the decision support system for initial analysis and treatment of retinal disease. This article develops a new Grasshopper Optimization with Fuzzy Edge Detection based Retinal Blood Vessel Segmentation and Classification (GOFED-RBVSC) model. The proposed GOFED-RBVSC model initially employs contrast enhancement process. Besides, GOAFED approach is employed to detect the edges in the retinal fundus images in which the use of GOA adjusts the membership functions. The ORB (Oriented FAST and Rotated BRIEF) feature extractor is exploited to generate feature vectors. Finally, Improved Conditional Variational Auto Encoder (ICAVE) is utilized for retinal image classification, shows the novelty of the work. The performance validation of the GOFED-RBVSC model is tested using benchmark dataset, and the comparative study highlighted the betterment of the GOFED-RBVSC model over the recent approaches.     

Automated Chinese Essay Scoring Based on Deep Learning

Writing is an important part of language learning and is considered the best approach to demonstrate the comprehensive language skills of students. Manually grading student essays is a time-consuming task; however, it is necessary. An automated essay scoring system can not only greatly improve the efficiency of essay scoring, but also provide more objective score. Therefore, many researchers have been exploring automated essay scoring techniques and tools. However, the technique of scoring Chinese essays is still limited, and its accuracy needs to be enhanced further. To improve the accuracy of the scoring model for a Chinese essay, we propose an automated scoring approach based on a deep learning model and validate its effect by conducting two comparison experiments. The experimental results indicate that the accuracy of the proposed model is significantly higher than that of multiple linear regression (MLR), which was commonly used in the past. The three accuracy rates of the proposed model are comparable to those of the novice teacher. The root mean square error (RMSE) of the proposed model is slightly lower than that of the novice teacher, and the correlation coefficient of the proposed model is also significantly higher than that of the novice teacher. Besides, when the predicted scores are not very low or very high, the two predicted models are as good as a novice teacher. However, when the predicted score is very high or very low, the results should be treated with caution.     

Review of Text Classification Methods on Deep Learning

Text classification has always been an increasingly crucial topic in natural language processing. Traditional text classification methods based on machine learning have many disadvantages such as dimension explosion, data sparsity, limited generalization ability and so on. Based on deep learning text classification, this paper presents an extensive study on the text classification models including Convolutional Neural Network-Based (CNN-Based), Recurrent Neural Network-Based (RNN-based), Attention Mechanisms-Based and so on. Many studies have proved that text classification methods based on deep learning outperform the traditional methods when processing large-scale and complex datasets. The main reasons are text classification methods based on deep learning can avoid cumbersome feature extraction process and have higher prediction accuracy for a large set of unstructured data. In this paper, we also summarize the shortcomings of traditional text classification methods and introduce the text classification process based on deep learning including text preprocessing, distributed representation of text, text classification model construction based on deep learning and performance evaluation.     

Multi-Scale Image Segmentation Model for Fine-Grained Recognition of Zanthoxylum Rust

Zanthoxylum bungeanum Maxim, generally called prickly ash, is widely grown in China. Zanthoxylum rust is the main disease affecting the growth and quality of Zanthoxylum. Traditional method for recognizing the degree of infection of Zanthoxylum rust mainly rely on manual experience. Due to the complex colors and shapes of rust areas, the accuracy of manual recognition is low and difficult to be quantified. In recent years, the application of artificial intelligence technology in the agricultural field has gradually increased. In this paper, based on the DeepLabV2 model, we proposed a Zanthoxylum rust image segmentation model based on the FASPP module and enhanced features of rust areas. This paper constructed a fine-grained Zanthoxylum rust image dataset. In this dataset, the Zanthoxylum rust image was segmented and labeled according to leaves, spore piles, and brown lesions. The experimental results showed that the Zanthoxylum rust image segmentation method proposed in this paper was effective. The segmentation accuracy rates of leaves, spore piles and brown lesions reached 99.66%, 85.16% and 82.47% respectively. MPA reached 91.80%, and MIoU reached 84.99%. At the same time, the proposed image segmentation model also had good efficiency, which can process 22 images per minute. This article provides an intelligent method for efficiently and accurately recognizing the degree of infection of Zanthoxylum rust.     

基于深度学习的包装组件检测系统研究

目的 针对人工分拣组成的零件包装盒常常会出现缺少部分零件的问题,开发一套集训练、识别、分选于一体的智能分拣系统.方法 在设计过程中,提出一种基于深度学习的改进Yolov3算法,针对工业现场光照、业零件形状和质地等实际因素,对Yolo算法的训练和检测进行改进,通过对包装盒产品的一次拍摄,检测出画面中出现的预设物体,并与标准设置相比对,从而判断出该盒内产品是否有缺料、多料的情况,以此分选出合格与否的包装盒.结果 在物体摆放相互重叠不超过20％的情况下,物体检测的准确率为98.2％,召回率为99.5％.结论 通过文中提出的改进算法,设计的检测系统能够在复杂的工业现场环境下正常工作,并能对包装的完整性进行准确的检测.