A fused deep learning architecture for viewpoint classification of echocardiography

PurposeThe aim of this study was to analyse and review deep learning convolutional neural networks for detecting and diagnosing early-stage dental caries on periapical radiographs.Materials and MethodsIn order to conduct this review, the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines were followed. Studies published from 2015 to 2021 under the keywords (deep convolutional neural network) AND (caries), (deep learning caries) AND (convolutional neural network) AND (caries) were systematically reviewed.ResultsWhen dental caries is improperly diagnosed, the lesion may eventually invade the enamel, dentin, and pulp tissue, leading to loss of tooth function. Rapid and precise detection and diagnosis are vital for implementing appropriate prevention and treatment of dental caries. Radiography and intraoral images are considered to play a vital role in detecting dental caries; nevertheless, studies have shown that 20% of suspicious areas are mistakenly diagnosed as dental caries using this technique; hence, diagnosis via radiography alone without an objective assessment is inaccurate. Identifying caries with a deep convolutional neural network-based detector enables the operator to distinguish changes in the location and morphological features of dental caries lesions. Deep learning algorithms have broader and more profound layers and are continually being developed, remarkably enhancing their precision in detecting and segmenting objects.ConclusionClinical applications of deep learning convolutional neural networks in the dental field have shown significant accuracy in detecting and diagnosing dental caries, and these models hold promise in supporting dental practitioners to improve patient outcomes.

Accurate short-term traffic flow forecasting facilitates active traffic control and trip planning. Most existing traffic flow models fail to make full use of the temporal and spatial features of traffic data. This study proposes a short-term traffic flow prediction model based on a convolution neural network (CNN) deep learning framework. In the proposed framework, the optimal input data time lags and amounts of spatial data are determined by a spatio-temporal feature selection algorithm (STFSA), and selected spatio-temporal traffic flow features are extracted from actual data and converted into a two-dimensional matrix. The CNN then learns these features to construct a predictive model. The effectiveness of the proposed method is evaluated by comparing the forecast results with actual traffic data. Other existing models are also evaluated for comparison. The proposed method outperforms baseline models in terms of accuracy.

Early detection of cutaneous melanoma (CM) is crucial for patient survival, yet avoiding overdiagnosis remains essential. Differentiating CM from benign melanoma simulators (MelSim) is challenging due to overlapping features. Deep learning convolutional neural networks (DL-CNNs) have demonstrated dermatologist-level accuracy in identifying CM. We hypothesized that support from DL-CNN could increase dermatologists' accuracy in differentiating CM from MelSim. This cross-sectional reader study analyzed the diagnostic performance of a DL-CNN and 27 dermatologists for 200 skin lesions (100 CM, 100 MelSim). The DL-CNN assigned malignancy scores ranging from 0 to 1 (> 0.5 indicating malignancy). Dermatologists assessed lesions across three levels: (I) dermoscopy only, (II) full case information (dermoscopy, close-up images, metadata), and (III) full case information plus DL-CNN scores. Primary outcomes were sensitivity, specificity, and ROC-AUC of dermatologists with or without DL-CNN-support (level-II versus -III). The DL-CNN and dermatologists in level-II showed a comparable sensitivity (95 % CI) of 90.0 % (82.6-94.5 %) and 90.1 % (86.9-93.2 %, p = 0.153), respectively. However, the DL-CNN's specificity (67.0 % (57.3-75.4 %) versus 73.2 % (69.1-77.3 %)) and ROC-AUC (0.889 (0.845-0.932) versus 0.951 (0.920-0.982)) were significantly lower than for dermatologists (all p < 0.01). When dermatologists integrated DL-CNN predictions (level-III) their sensitivity increased to 91.4 % (88.3-94.5 %, p < 0.001) without markedly changing specificity (74.2 % (70.6-77.7 %, p = 0.435)) or ROC-AUC (0.954 (0.927-0.982, p = 0.581)). Collaboration with a DL-CNN slightly improved dermatologists' diagnostic accuracy in a mixed CM and MelSim dataset, by increasing sensitivity without a loss of specificity. The DL-CNN's level of sensitivity in this difficult-to-diagnose dataset underlines the potential as an assistant diagnostic tool.

The study was aimed to explore the effect of information health data based on deep learning of neural network to diagnose the infectious factors of patients with chronic glomerular disease (CGD) and evaluate its diagnostic effect. Ninety patients with CGD were selected and randomly rolled into control group A, control group B, and observation group, with 30 cases in each group. Big data scientific research analysis platform was used for data integration, convolutional neural network (CNN) was employed for feature analysis, correlation analysis, and screening of disease‐related biomarkers. The patients were diagnosed by observation of symptoms and signs, combined diagnosis of blood test and urine test, and information health data diagnosis based on deep learning CNN. As a result, the specificity, sensitivity, and accuracy of information health data diagnosis based on deep learning CNN were 78.9%, 87.6%, and 92.1%, respectively. The main sources of infections in patients were lung infections, bloodstream infections, urinary system infections, skin and soft tissue infections, and upper respiratory tract infections. Amongst them, lung infection accounted for the highest proportion, reaching 65.4%, followed by blood infection (11.2%) and skin tissue infection (9.6%). The pathogens of infection were mainly bacteria, viruses, fungi, tuberculosis, and pneumocystis pneumonia (PCP), amongst which bacterial infections accounted for the highest proportion (31.5%), followed by PCP (25.6%). In short, the information health data based on deep learning CNN had high specificity, sensitivity, and accuracy for the diagnosis of CGD. The main infectious factors of CGD were pulmonary infection and blood infection, and the pathogens were mainly bacteria and viruses.

As one of the representative algorithms of deep learning, a convolutional neural network (CNN) with the advantage of local perception and parameter sharing has been rapidly developed. CNN-based detection technology has been widely used in computer vision, natural language processing, and other fields. Fresh fruit production is an important socioeconomic activity, where CNN-based deep learning detection technology has been successfully applied to its important links. To the best of our knowledge, this review is the first on the whole production process of fresh fruit. We first introduced the network architecture and implementation principle of CNN and described the training process of a CNN-based deep learning model in detail. A large number of articles were investigated, which have made breakthroughs in response to challenges using CNN-based deep learning detection technology in important links of fresh fruit production including fruit flower detection, fruit detection, fruit harvesting, and fruit grading. Object detection based on CNN deep learning was elaborated from data acquisition to model training, and different detection methods based on CNN deep learning were compared in each link of the fresh fruit production. The investigation results of this review show that improved CNN deep learning models can give full play to detection potential by combining with the characteristics of each link of fruit production. The investigation results also imply that CNN-based detection may penetrate the challenges created by environmental issues, new area exploration, and multiple task execution of fresh fruit production in the future.

Auto-detection and segmentation of involved lymph nodes in HPV-associated oropharyngeal cancer using a convolutional deep learning neural network

Breast cancer patients who have pathological complete response (pCR) to neoadjuvant chemotherapy (NAC) are more likely to have better clinical outcomes. The ability to predict which patient will respond to NAC early in the treatment course is important because it could help to minimize unnecessary toxic NAC and to modify regimens mid-treatment to achieve better efficacy. Machine learning (ML) is increasingly being used in radiology and medicine because it can identify relationships amongst complex data elements to inform outcomes without the need to specify such relationships a priori. One of the most popular deep learning methods that applies to medical images is the Convolutional Neural Networks (CNN). In contrast to supervised ML, deep learning CNN can operate on the whole images without requiring radiologists to manually contour the tumor on images. Although there have been many review papers on supervised ML prediction of pCR, review papers on deep learning prediction of pCR are sparse. Deep learning CNN could also incorporate multiple image types, clinical data such as demographics and molecular subtypes, as well as data from multiple treatment time points to predict pCR. The goal of this study is to perform a systematic review of deep learning methods that use whole-breast MRI images without annotation or tumor segmentation to predict pCR in breast cancer.

Application of Deep Learning Convolutional Neural Networks for Internal Tablet Defect Detection: High Accuracy, Throughput, and Adaptability

Introduction Catathrenia is a rare sleep-related breathing disorder characterized by recurrent monotonous groaning during sleep. The acoustic characteristics of groaning and snoring sounds have been investigated. This study aims to propose a deep convolutional neural network (CNN) for automatic binary classification. Methods This study consisted of 3728 episodes of groaning sounds and 4577 episodes of snoring sounds obtained from synchronized audio of full-night polysomnography. Four features extracted from log-scaled mel-spectrograms were used as input. The background gaussian noise and the time-shifting were used to augment the dataset. The dataset was randomly split into training (70%), validation (15%), and testing datasets (15%). A deep learning convolutional neural network architecture was trained and evaluated. Results The proposed CNN model achieves an accuracy of 95.0% on the binary classification of groaning and snoring sounds. The model attains a sensitivity/recall of 96.4%, a specificity of 93.4%, and an F1 score of 94.84%. Conclusion The proposed CNN architecture has performed well in the automatic binary classification of groaning and snoring sounds, which could reduce difficulties in acoustic analyses of groaning episode detection. The model needs to be verified with more audio data before it can be put into clinical use better. Support (if any)

Accurate and reliable detection of white matter hyperintensities and their volume quantification can provide valuable clinical information to assess neurologic disease progression. In this work, a stacked generalization ensemble of orthogonal 3D convolutional neural networks, StackGen-Net, is explored for improving automated detection of white matter hyperintensities in 3D T2-FLAIR images. Individual convolutional neural networks in StackGen-Net were trained on 2.5D patches from orthogonal reformatting of 3D-FLAIR (n = 21) to yield white matter hyperintensity posteriors. A meta convolutional neural network was trained to learn the functional mapping from orthogonal white matter hyperintensity posteriors to the final white matter hyperintensity prediction. The impact of training data and architecture choices on white matter hyperintensity segmentation performance was systematically evaluated on a test cohort (n = 9). The segmentation performance of StackGen-Net was compared with state-of-the-art convolutional neural network techniques on an independent test cohort from the Alzheimer's Disease Neuroimaging Initiative-3 (n = 20). StackGen-Net outperformed individual convolutional neural networks in the ensemble and their combination using averaging or majority voting. In a comparison with state-of-the-art white matter hyperintensity segmentation techniques, StackGen-Net achieved a significantly higher Dice score (0.76 [SD, 0.08], F1-lesion (0.74 [SD, 0.13]), and area under precision-recall curve (0.84 [SD, 0.09]), and the lowest absolute volume difference (13.3% [SD, 9.1%]). StackGen-Net performance in Dice scores (median = 0.74) did not significantly differ (P = .22) from interobserver (median = 0.73) variability between 2 experienced neuroradiologists. We found no significant difference (P = .15) in white matter hyperintensity lesion volumes from StackGen-Net predictions and ground truth annotations. A stacked generalization of convolutional neural networks, utilizing multiplanar lesion information using 2.5D spatial context, greatly improved the segmentation performance of StackGen-Net compared with traditional ensemble techniques and some state-of-the-art deep learning models for 3D-FLAIR.

The assessment of bone age has applications in a wide variety of fields: from orthodontics to immigration. The traditional non-automated methods are time-consuming and subject to inter- and intra-observer variability. This is the first study of its kind done on the Indian population. In this study, we analyse different pre-processing techniques and architectures to determine the degree of maturation (i.e. cervical vertebral maturation [CVM]) from cephalometric radiographs using machine learning algorithms. Cephalometric radiographs-labelled with the correct CVM stage using Baccetti et al. method-from 383 individuals aged between 10 and 36 years were used in the study. Data expansion and in-place data augmentation were used to handle high data imbalances. Different pre-processing techniques like Sobel filters and canny edge detectors were employed. Several deep learning convolutional neural network (CNN) architectures along with numerous pre-trained models like ResNet-50 and VGG-19 were analysed for their efficacy on the dataset. Models with 6 and 8 convolutional layers trained on 64 × 64-size grayscale images trained the fastest and achieved the highest accuracy of 94%. Pre-trained ResNet-50 with the first 49 layers frozen and VGG-19 with 10 layers frozen to training had remarkable performances on the dataset with accuracies of 91% and 89%, respectively. Custom deep CNN models with 6-8 layers on 64 × 64-sized greyscale images were successfully used to achieve high accuracies to classify the majority classes. This study is a launchpad in the development of an automated method for bone age assessment from lateral cephalograms for clinical purposes.

Medulloblastoma (MB) is a dangerous malignant pediatric brain tumor that could lead to death. It is considered the most common pediatric cancerous brain tumor. Precise and timely diagnosis of pediatric MB and its four subtypes (defined by the World Health Organization (WHO)) is essential to decide the appropriate follow-up plan and suitable treatments to prevent its progression and reduce mortality rates. Histopathology is the gold standard modality for the diagnosis of MB and its subtypes, but manual diagnosis via a pathologist is very complicated, needs excessive time, and is subjective to the pathologists’ expertise and skills, which may lead to variability in the diagnosis or misdiagnosis. The main purpose of the paper is to propose a time-efficient and reliable computer-aided diagnosis (CADx), namely MB-AI-His, for the automatic diagnosis of pediatric MB and its subtypes from histopathological images. The main challenge in this work is the lack of datasets available for the diagnosis of pediatric MB and its four subtypes and the limited related work. Related studies are based on either textural analysis or deep learning (DL) feature extraction methods. These studies used individual features to perform the classification task. However, MB-AI-His combines the benefits of DL techniques and textural analysis feature extraction methods through a cascaded manner. First, it uses three DL convolutional neural networks (CNNs), including DenseNet-201, MobileNet, and ResNet-50 CNNs to extract spatial DL features. Next, it extracts time-frequency features from the spatial DL features based on the discrete wavelet transform (DWT), which is a textural analysis method. Finally, MB-AI-His fuses the three spatial-time-frequency features generated from the three CNNs and DWT using the discrete cosine transform (DCT) and principal component analysis (PCA) to produce a time-efficient CADx system. MB-AI-His merges the privileges of different CNN architectures. MB-AI-His has a binary classification level for classifying among normal and abnormal MB images, and a multi-classification level to classify among the four subtypes of MB. The results of MB-AI-His show that it is accurate and reliable for both the binary and multi-class classification levels. It is also a time-efficient system as both the PCA and DCT methods have efficiently reduced the training execution time. The performance of MB-AI-His is compared with related CADx systems, and the comparison verified the powerfulness of MB-AI-His and its outperforming results. Therefore, it can support pathologists in the accurate and reliable diagnosis of MB and its subtypes from histopathological images. It can also reduce the time and cost of the diagnosis procedure which will correspondingly lead to lower death rates.

Deep Learning Facilitates Automation of Wall Layer Quantification in Heart Transplant Coronary OCT

It aimed to analyze the value of deep learning algorithm combined with magnetic resonance imaging (MRI) in the risk diagnosis and prognosis of endometrial cancer (EC). Based on the deep learning convolutional neural network (CNN) architecture residual network with 101 layers (ResNet-101), spatial attention and channel attention modules were introduced to optimize the model. A retrospective collection of MRI image data from 210 EC patients was used for model segmentation and reconstruction, with 140 cases as the test set and 70 cases as the validation set. The performance was compared with traditional ResNet-101 model, ResNet-101 model based on spatial attention mechanism (SA-ResNet-101), and ResNet-101 model based on channel attention mechanism (CA-ResNet-101), using accuracy (AC), precision (PR), recall (RE), and F1 score as evaluation metrics. Among the 70 cases in the validation set, there were 45 cases of low-risk EC and 25 cases of high-risk EC. Using ROC curve analysis, it was found that the area under the curve (AUC) for the diagnosis of high-risk EC of the proposed model in this article (0.918) was visibly larger as against traditional ResNet-101 (0.613), SA-ResNet-101 (0.760), and CA-ResNet-101 models (0.758). The AC, PR, RE, and F1 values of the proposed model for the diagnosis of EC risk were visibly higher (P < 0.05). In the validation set, postoperative recurrence occurred in 13 cases and did not occur in 57 cases. Using ROC curve analysis, it was found that the AUC for postoperative recurrence prediction of the patients by the proposed model (0.926) was visibly larger as against traditional ResNet-101 (0.620), SA-ResNet-101 (0.729), and CA-ResNet-101 models (0.767). The AC, PR, RE, and F1 values of the proposed model for postoperative recurrence prediction were visibly higher (P < 0.05). The proposed model in this article, assisted by MRI, presented superior performance in diagnosing high-risk EC patients, with higher sensitivity (Sen) and specificity (Spe), and also demonstrated excellent predictive AC in postoperative recurrence prediction.

With the aim of identifying possible mechanical faults in unmanned aerial vehicle (UAV) rotors during operation, this paper proposes a method based on interval sampling reconstruction of vibration signals and one-dimensional convolutional neural network (1D-CNN) deep learning. Firstly, experiments were designed to collect the vibration acceleration signals of a UAV working at high speed under three states (normal, rotor damage by varying degrees, and rotor crack by different degrees). Then, considering the powerful feature extraction and complex data analysis abilities of 1D-CNN, an effective deep learning model for fault identification is established utilizing 1D-CNN. During analysis, it is found that the recognition rate for minor faults is not ideal, with all weak states being identified as normal, reducing the overall identification accuracy, when using conventional sequential sampling to construct learning sample sets. To this end, in order to make the sample data cover the whole process of data collection as much as possible, a learning sample processing method based on interval sampling reconstruction of the vibration signal is proposed. And it is also verified that the reconstructed sample set can easily reflect the global information of mechanical operation. Finally, according to the comparison of analysis results, the recognition rate of the deep learning model for different degrees of faults is greatly improved, and minor faults could also be accurately identified through this method. The results show that the 1D-CNN deep learning model could diagnose and identify UAV rotor damage faults accurately, by incorporating the proposed method of interval sampling reconstruction.