Detection of pulmonary micronodules in computed tomography images and false positive reduction using 3D convolutional neural networks

Manual detection of small uncalcified pulmonary nodules (diameter <4 mm) in thoracic computed tomography (CT) scans is a tedious and error-prone task. Automatic detection of disperse micronodules is, thus, highly desirable for improved characterization of the fatal and incurable occupational pulmonary diseases. Here, we present a novel computer-assisted detection (CAD) scheme specifically dedicated to detect micronodules. The proposed scheme consists of a candidate-screening module and a false positive (FP) reduction module. The candidate-screening module is initiated by a lung segmentation algorithm and is followed by a combination of 2D/3D features-based thresholding parameters to identify plausible micronodules. The FP reduction module employs a 3D convolutional neural network (CNN) to classify each identified candidate. It automatically encodes the discriminative representations by exploiting the volumetric information of each candidate. A set of 872 micro-nodules in 598 CT scans marked by at least two radiologists are extracted from the Lung Image Database Consortium and Image Database Resource Initiative to test our CAD scheme. The CAD scheme achieves a detection sensitivity of 86.7% (756/872) with only 8 FPs/scan and an AUC of 0.98. Our proposed CAD scheme efficiently identifies micronodules in thoracic scans with only a small number of FPs. Our experimental results provide evidence that the automatically generated features by the 3D CNN are highly discriminant, thus making it a well-suited FP reduction module of a CAD scheme.     

A serialized classification method for pulmonary nodules based on lightweight cascaded convolutional neural network-long short-term memory

Computer Assisted Diagnosis (CAD) is an effective method to detect lung cancer from computed tomography (CT) scans. The development of artificial neural network makes CAD more accurate in detecting pathological changes. Due to the complexity of the lung environment, the existing neural network training still requires large datasets, excessive time, and memory space. To meet the challenge, we analysis 3D volumes as serialized 2D slices and present a new neural network structure lightweight convolutional neural network (CNN)-long short-term memory (LSTM) for lung nodule classification. Our network contains two main components: (a) optimized lightweight CNN layers with tiny parameter space for extracting visual features of serialized 2D images, and (b) LSTM network for learning relevant information among 2D images. In all experiments, we compared the training results of several models and our model achieved an accuracy of 91.78% for lung nodule classification with an AUC of 93%. We used fewer samples and memory space to train the model, and we achieved faster convergence. Finally, we analyzed and discussed the feasibility of migrating this framework to mobile devices. The framework can also be applied to cope with the small amount of training data and the development of mobile health device in future.     

基于迁移学习与深度残差网络的滚动轴承快速故障诊断算法EI北大核心CSCD

针对现有基于深度学习的滚动轴承故障诊断算法训练参数量大,训练时间长且需要大量训练样本的缺点,提出了一种基于迁移学习(TL)与深度残差网络(ResNet)的快速故障诊断算法(TL-ResNet)。首先开发了一种将短时傅里叶变换(STFT)与伪彩色处理相结合的振动信号转三通道图像数据的方法;然后将在ImageNet数据集上训练的ResNet18模型作为预训练模型,通过迁移学习的方法,应用到滚动轴承故障诊断领域当中;最后对滚动轴承在不同工况下的故障诊断问题,提出了采用小样本迁移的方法进行诊断。在凯斯西储大学(CWRU)与帕德博恩大学(PU)数据集上进行了试验,TL-ResNet的诊断准确率分别为99.8%与95.2%,且在CWRU数据集上TL-ResNet的训练时间仅要1.5 s,这表明本算法优于其他的基于深度学习的故障诊断算法与经典算法,可用于实际工业环境中的快速故障诊断。     

Computer‐aided diagnosis system for the detection of bronchiectasis in chest computed tomography images

A computer‐aided diagnosis (CAD) system has been developed for the detection of bronchiectasis from computed tomography (CT) images of chest. A set of CT images of the chest with known diagnosis were collected and these images were first denoised using Wiener filter. The lung tissue was then segmented using optimal thresholding. The Pathology Bearing Regions (PBRs) were then extracted by applying pixel‐based segmentation. For each PBR, a gray level co‐occurrence matrix (GLCM) was constructed. From the GLCM texture features were extracted and feature vectors were constructed. A probabilistic neural network (PNN) was constructed and trained using this set of feature vectors. The images together with the PBRs and the corresponding feature vector and diagnosis were stored in an image database. Rules for diagnosis and for determining the severity of the disease were generated by analyzing the images known to be affected by bronchiectasis. The rules were then validated by a human expert. The validated rules were stored in the Knowledge Base. When a physician gives a CT image to the CAD system, it first transforms the image into a set of feature vectors, one for each PBR in the image. It then performs the diagnosis using two techniques: PNN and mahalanobis distance measure. The final diagnosis and the severity of the disease are determined by correlating the diagnosis determined by both the techniques in consultation with the knowledge base. The system also retrieves similar cases from the database. Thus, this system would aid the physicians in diagnosing bronchiectasis. © 2009 Wiley Periodicals, Inc. Int J Imaging Syst Technol, 19, 290–298, 2009     

基于深度神经网络的水声FBMC通信信号检测方法

针对传统水声滤波器组多载波(Filter Bank Multi-Carrier,FBMC)通信接收端需经过信道估计和均衡才可恢复出发送符号,系统复杂度高且信道估计精度不佳等问题.文章将深度神经网络融入到水声多载波通信当中,提出一种基于深度神经网络的水声FBMC信号检测方法.在训练阶段通过大量的数据迭代、调试超参数和优化...     

Transfer Learning-based Computer-aided Diagnosis System for Predicting Grades of Diabetic Retinopathy

Diabetic retinopathy (DR) diagnosis through digital fundus images requires clinical experts to recognize the presence and importance of many intricate features. This task is very difficult for ophthalmologists and time-consuming. Therefore, many computer-aided diagnosis (CAD) systems were developed to automate this screening process of DR. In this paper, a CAD-DR system is proposed based on preprocessing and a pre-train transfer learning-based convolutional neural network (PCNN) to recognize the five stages of DR through retinal fundus images. To develop this CAD-DR system, a preprocessing step is performed in a perceptual-oriented color space to enhance the DR-related lesions and then a standard pre-train PCNN model is improved to get high classification results. The architecture of the PCNN model is based on three main phases. Firstly, the training process of the proposed PCNN is accomplished by using the expected gradient length (EGL) to decrease the image labeling efforts during the training of the CNN model. Secondly, the most informative patches and images were automatically selected using a few pieces of training labeled samples. Thirdly, the PCNN method generated useful masks for prognostication and identified regions of interest. Fourthly, the DR-related lesions involved in the classification task such as micro-aneurysms, hemorrhages, and exudates were detected and then used for recognition of DR. The PCNN model is pre-trained using a high-end graphical processor unit (GPU) on the publicly available Kaggle benchmark. The obtained results demonstrate that the CAD-DR system outperforms compared to other state-of-the-art in terms of sensitivity (SE), specificity (SP), and accuracy (ACC). On the test set of 30,000 images, the CAD-DR system achieved an average SE of 93.20%, SP of 96.10%, and ACC of 98%. This result indicates that the proposed CAD-DR system is appropriate for the screening of the severity-level of DR.     

A Novel Technique for Early Detection of COVID-19

COVID-19 is a global pandemic disease, which results from a dangerous coronavirus attack, and spreads aggressively through close contacts with infected people and artifacts. So far, there is not any prescribed line of treatment for COVID-19 patients. Measures to control the disease are very limited, partly due to the lack of knowledge about technologies which could be effectively used for early detection and control the disease. Early detection of positive cases is critical in preventing further spread, achieving the herd immunity, and saving lives. Unfortunately, so far we do not have effective toolkits to diagnose very early detection of the disease. Recent research findings have suggested that radiology images, such as X-rays, contain significant information to detect the presence of COVID-19 virus in early stages. However, to detect the presence of the disease in in very early stages from the X-ray images by the naked eye is not possible. Artificial Intelligence (AI) techniques, machine learning in particular, are known to be very helpful in accurately diagnosing many diseases from radiology images. This paper proposes an automatic technique to classify COVID-19 patients from their computerized tomography (CT) scan images. The technique is known as Advanced Inception based Recurrent Residual Convolution Neural Network (AIRRCNN), which uses machine learning techniques for classifying data. We focus on the Advanced Inception based Recurrent Residual Convolution Neural Network, because we do not find it being used in the literature. Also, we conduct principal component analysis, which is used for dimensional deduction. Experimental results of our method have demonstrated an accuracy of about 99%, which is regarded to be very efficient.     

Geometrical approach for automatic detection of liquid surfaces in 3D computed tomography baggage imagery

This study presents a novel method for liquid detection within three-dimensional (3D) computed tomography (CT) baggage inspection imagery. Liquid detection within airport security is currently of significant interest due to security threats associated with liquid explosives. In this paper, we propose a robust technique based on the automatic identification of universal geometric properties of liquids within 3D space. The proposed approach is based on two stages of geometric fitting. First, we identify the 3D plane which fits to the horizontally oriented surface of the liquid recognising the universal self-levelling property of liquids in any given container. Second, we conduct two-dimensional shape analysis to highlight the shape of the liquid surface at a given level within the container using a least squares elliptical fitting approach. The proposed approach relies on the fact that occurrences of such perfectly aligned horizontal planes within a 3D CT security baggage scan are generally unlikely. Occurrences of such instance are thus indicative of liquid presence. Our results, over an extended set of complex test examples, confirm a liquid detection rate of 85–98% with a moderate processing time. Furthermore, as this proposed approach is based purely on the geometric properties of liquids and robust geometrical shape detection, this methodology is intrinsic to the 3D nature of the resulting CT data and not dependent on any exemplar training imagery.     

A novel hybrid artificial neural network technique for the early skin cancer diagnosis using color space conversions of original images

In this study, an innovative hybrid machine learning-technique is used for the early skin cancer diagnosis fusing Convolutional Neural Network and Multilayer Perceptron to analyze images and information related to the skin cancer. This information is extracted manually after applying different color space conversions on the original images for better screening of the lesions. The proposed architecture is compared with standalone architecture in addition to some other techniques by commonly used evaluation metrics. HAM10000 dataset is used for training and testing as this data contain seven different skin lesions. The novelty of the proposed hybrid model is the structure of the network which handles structured data (patients' metadata and other useful features from different color spaces related to the illumination, energy, darkness, etc.) and unstructured data (images). The results show an overall 86%, 95% top-1 and top-2 accuracy respectively, and 96% area under the curve for the seven classes. The study demonstrates the superiority of the proposed hybrid model with a 2% improvement in the accuracy over the standalone model and a promising behavior as compared to the ensemble techniques. The follow-up research will include more patient data to develop a skin cancer detection device.