Vehicle Classification Using Electromagnetic Leakage Signals Based on Feature Fusion and Residual CNN

BackgroundP53-mutated hepatocellular carcinoma (HCC) is an aggressive subtype with poor prognosis, currently diagnosed only by invasive biopsy. Noninvasive preoperative prediction could guide personalized treatment. This study aimed to develop and evaluate a preoperative method for predicting P53-mutated HCC using deep learning (DL) and radiomics model based on magnetic resonance imaging (MRI).MethodsIn this retrospective study, we included 320 consecutive patients who underwent surgical resection for HCC between January 2020 and February 2021, had postoperative P53 immunohistochemistry results, and showed no evidence of extrahepatic metastases preoperatively. Patients were randomly assigned to a Training cohort (n=224) and a Validation cohort (n=96). Clinical risk factors were identified through stepwise regression analysis, and a clinical prediction model was built. We developed a radiomics model and multiple convolutional neural network (CNN) models using dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) of HCC. Additionally, clinical models were constructed based on clinical and imaging features. Two fusion strategies were employed to build DL fusion models for the arterial phase (AP), portal venous phase (PVP), and delayed phase (DP): the multi-channel (MC) DL model and the feature fusion (FF) DL model. Ultimately, we integrated the optimal DL model with the radiomics and clinical models into a combined model, developed nomograms to predict P53-mutated HCC, and evaluated its performance using receiver operating characteristic (ROC) analysis, calibration, and decision curve analysis.ResultsOf 320 patients, 259 were men and 61 were women (mean age, 58±11 years). There were 183 cases of P53-mutated HCC and 137 cases of non-P53-mutated HCC. In the MC DL model, ResNet34 achieved area under the curve (AUC) values of 0.731 [95% confidence interval (CI): 0.654–0.808] in the training cohort (TC) and 0.652 (95% CI: 0.447–0.858) in the validation cohort (VC), outperforming other CNNs. In the FF DL model, ResNet101 demonstrated an AUC of 0.779 (95% CI: 0.719–0.839) for TC and 0.663 (95% CI: 0.552–0.774) for VC, showing superior performance compared to other CNNs. Integration of the clinical, radiomics, and DL models improved the AUC to 0.838 (95% CI: 0.786–0.891) in the TC and 0.702 (95% CI: 0.598–0.807) in the VC.ConclusionsThe combination of clinical, radiomics, and DL models provides an effective tool for preoperative prediction of P53 mutation status in patients with HCC.

Failure to predict stroke promptly may lead to delayed treatment, causing severe consequences like permanent neurological damage or death. Early detection using deep learning (DL) and machine learning (ML) models can enhance patient outcomes and mitigate the long-term effects of strokes. The aim of this study is to compare these models, exploring their efficacy in predicting stroke. This study analyzed a dataset comprising 663 records from patients hospitalized at Hazrat Rasool Akram Hospital in Tehran, Iran, including 401 healthy individuals and 262 stroke patients. A total of eight established ML (SVM, XGB, KNN, RF) and DL (DNN, FNN, LSTM, CNN) models were utilized to predict stroke. Techniques such as 10-fold cross-validation and hyperparameter tuning were implemented to prevent overfitting. The study also focused on interpretability through Shapley Additive Explanations (SHAP). The evaluation of model’s performance was based on accuracy, specificity, sensitivity, F1-score, and ROC curve metrics. Among DL models, LSTM showed superior sensitivity at 96.15%, while FNN exhibited better specificity (96.0%), accuracy (96.0%), F1-score (95.0%), and ROC (98.0%) among DL models. For ML models, RF displayed higher sensitivity (99.9%), accuracy (99.0%), specificity (100%), F1-score (99.0%), and ROC (99.9%). Overall, RF outperformed all models, while DL models surpassed ML models in most metrics except for RF. DL models (CNN, LSTM, DNN, FNN) achieved sensitivities from 93.0 to 96.15%, specificities from 80.0 to 96.0%, accuracies from 92.0 to 96.0%, F1-scores from 87.34 to 95.0%, and ROC scores from 95.0 to 98.0%. In contrast, ML models (KNN, XGB, SVM) showed sensitivities between 29.0% and 94.0%, specificities between 89.47% and 96.0%, accuracies between 71.0% and 95.0%, F1-scores between 44.0% and 95.0%, and ROC scores between 64.0% and 95.0%. This study demonstrates the efficacy of DL and ML models in predicting stroke, with the RF models outperforming all others in key metrics. While DL models generally surpassed ML models, RF’s exceptional performance highlights the potential of combining these technologies for early stroke detection, significantly improving patient outcomes by preventing severe consequences like permanent neurological damage or death.

Speaker identification systems perform almost ideally in neutral talking environments. However, these systems perform poorly in stressful talking environments. In this paper, we present an effective approach for enhancing the performance of speaker identification in stressful talking environments based on a novel radial basis function neural network-convolutional neural network (RBFNN-CNN) model. In this research, we applied our approach to two distinct speech databases: a local Arabic Emirati-accent dataset and a global English Speech Under Simulated and Actual Stress (SUSAS) corpus. To the best of our knowledge, this is the first work that addresses the use of an RBFNN-CNN model in speaker identification under stressful talking environments. Our speech identification models select the finest speech signal representation through the use of Mel-frequency cepstral coefficients (MFCCs) as a feature extraction method. A comparison among traditional classifiers such as support vector machine (SVM), multilayer perceptron (MLP), k-nearest neighbors algorithm (KNN) and deep learning models, such as convolutional neural network (CNN) and recurrent neural network (RNN), was conducted. The results of our experiments show that speaker identification performance in stressful environments based on the RBFNN-CNN model is higher than that with the classical and deep machine learning models.

Background Interstitial/Alveolar Syndrome (IS) is a condition detectable on lung ultrasound (LUS) that indicates underlying pulmonary or cardiac diseases associated with significant morbidity and increased mortality rates. The diagnosis of IS using LUS can be challenging and time-consuming, and it requires clinical expertise. Methods In this study, multiple Convolutional Neural Network (CNN) deep learning (DL) models were trained, acting as binary classifiers, to accurately screen for IS from LUS frames by differentiating between IS-present and healthy cases. The CNN DL models were initially pre-trained using a generic image dataset to learn general visual features (ImageNet), and then fine-tuned on our specific dataset of 108 LUS clips from 54 patients (27 healthy and 27 with IS), with two clips per patient, to perform a binary classification task. Each frame within a clip was assessed to determine the presence of IS features or to confirm a healthy lung status. The dataset was split into training (70%), validation (15%), and testing (15%) sets. Following the process of fine-tuning, we successfully extracted features from pre-trained DL models. These extracted features were utilised to train multiple machine learning (ML) classifiers, hence the trained ML classifiers yielded significantly improved accuracy in IS classification. Advanced visual interpretation techniques, such as heatmaps based on Gradient-weighted Class Activation Mapping (Grad-CAM) and Local Interpretable Model-Agnostic explanations (LIME), were implemented to further analyse the outcomes. Results The best-trained ML model achieved a test accuracy of 98.2%, with specificity, recall, precision, and F1-score values all above 97.9%. Our study demonstrates, for the first time, the feasibility of using a pre-trained CNN with the feature extraction and fusion technique as a diagnostic tool for IS screening on LUS frames, providing a time-efficient and practical approach to clinical decision-making. Conclusion This study confirms the practicality of using pre-trained CNN models, with the feature extraction and fusion technique, for screening IS through LUS frames. This represents a noteworthy advancement in improving the efficiency of diagnosis. In the next steps, validation on larger datasets will assess the applicability and robustness of these CNN models in more complex clinical settings.

In the field of underwater acoustic recognition, machine learning methods rely on a large number of datasets to achieve high accuracy, while the actual collected signal samples are often very scarce, which has a great impact on the recognition performance. This paper presents a recognition method of an underwater acoustic target by the data augmentation technique and the residual convolutional neural network (CNN) model, which is used to expand training samples to improve recognition performance. As a representative model in residual CNN, the ResNet18 model is used for recognition. The whole process mainly includes mel-frequency cepstral coefficient (MFCC) feature extraction, data augmentation processing, and ResNet18 model recognition. On the base of the traditional data augmentation, this study used the deep convolutional generative adversarial network (DCGAN) model to realize the expansion of underwater acoustic samples and compared the recognition performance of support vector machine (SVM), common CNN, VGG19, and ResNet18. The recognition results of the MFCC, constant Q transform (CQT), and low-frequency analyzer and recorder (LOFAR) spectrum were also analyzed and compared. Experimental results showed that the recognition accuracy of the MFCC feature was better than that of other features at the same method, and using the data augmentation method could obviously improve the recognition performance. Moreover, the recognition performance of ResNet18 using data enhancement technology was better than that of other models, which was due to the combination of the data expansion advantage of data augmentation technology and the deep feature extracting ability of the residual CNN model. In addition, although this method was used for ship recognition in this paper, it is not limited to this. This method is also applicable to other target voice recognition, such as natural sound and underwater voice biometrics.

LSTM, WaveNet, and 2D CNN for nonlinear time history prediction of seismic responses

Distracting while driving is a serious issue that causes serious direct and indirect harm to the society. To avoid these problems, detecting dangerous drivers’ behaviour is very important.This research focuses on detecting driver behaviour with a combination of artificial deep learning and machine learning models with genetic algorithm. Most of the previous works have focused on using convolutional neural network as deep learning model or support vector machine as machine learning model for actions detection of drivers from input images. The proposed structure will use genetic algorithms to first choose the structure of feature extractor from famous CNN models such as VGG19, ResNet50, and DenseNet121. After mentioning feature extractor, proposed framework contains two layer of dense layer for classification as a deep learning model. On the machine learning side K nearest neighbor, random forest, support vector machine, and extreme boost algorithms have been used as classifiers. Genetic algorithms will specify number of neurons and activation functions of these layers for deep learning and hyperparameters such as number of estimators for machine learning models. Proposed model has been developed with the use of state farm dataset that contains information of 1 safe driving class and 9 dangerous behaviours such as texting while driving, talking with passengers, drinking, etc. Experimental results indicate 99.80% accuracy for classification of the state farm distracted driver detection with combination of genetic algorithms and deep neural networks. Compared to similar research, the proposed approach has shown superior results for classification of state farm distracted driver detection. Proposed approach chooses the feature extraction model and hyperparameters of the classification layer automatically. Thus it can be used for driving behaviour classification with seeing new situation too. Proposed framework can be used as a real time driver’s distraction detection to decrease car traffic accidents and alleviate corresponding damages to the drivers.

A novel deep-learning technique for forecasting oil price volatility using historical prices of five precious metals in context of green financing – A comparison of deep learning, machine learning, and statistical models

Oranges, mandarins, bitter oranges, and lemons are examples of citrus fruits that make delicious meals and are highly nutritious. Citrus fruits suffer from a variety of infections that affect their yield. The Department of Agriculture wants to increase the production of oranges and lemons. On the other hand, several plant diseases and their advanced stages have impacted production. The quality of fruit influences market value and its financial effect. Therefore, accurate detection of ailments and their severity is crucial for improving the output and market value of oranges and lemons. To automatically evaluate and predict diseases in citrus leaves and fruits, this paper has proposed a modified convolutional neural network (ICNN) model. Python is used to create the ICNN model, and testing is performed using benchmark datasets from various repositories. The research presented here shows that ICNN performs better than traditional deep learning and machine learning models, such as the Convolutional Neural Network (CNN) and K-Nearest Neighbours (KNN). This illustrates how machine learning models require supplementary approaches to extract parameters from data that arrives in non-automated ways. Additionally, to improve the accuracy of their classification or prediction, deep learning models require pre-trained models. As a result, ICNN, an enhanced deep learning model that can automatically predict disease with higher accuracy than other models, represents an advancement over standard CNNs. Compared with KNN and CNN, ICNN achieves 99.69% accuracy.

Owing to the 4.0 industrial revolution condition monitoring maintenance is widely accepted as a useful approach to avoiding plant disturbances and shutdown. Recently, Motor Current Signature Analysis (MCSA) is widely reported as a condition monitoring technique in the detection and identification of individual and multiple Induction Motor (IM) faults. However, checking the fault detection and classification with deep learning models and its comparison among themselves or conventional approaches is rarely reported in the literature. Therefore, in this work, we present the detection and identification of induction motor faults with MCSA and three Deep Learning (DL) models namely MLP, LSTM, and 1D-CNN. Initially, we have developed the model of Squirrel Cage induction motor in MATLAB and simulated it for single phasing and stator winding faults (SWF) using Fast Fourier Transform (FFT), Short Time Fourier Transform (STFT), and Continuous Wavelet Transform (CWT) to detect and identify the healthy and unhealthy conditions with phase to ground, single phasing and in multiple fault conditions using Motor Current Signature Analysis. The faults impact on stator current is presented in the time and frequency domain (i.e., power spectrum). The simulation results show that the scalogram has shown good results in time-frequency analysis for fault and showing its impact on the energy of current during individual fault and multiple fault conditions. This is further investigated with three deep learning models (i.e., MLP, LSTM, and 1D-CNN) for checking the fault detection and identification (i.e., classification) improvement in a three-phase induction motor. By simulating the three-phase induction motor in various healthy and unhealthy conditions in MATLAB, we have collected current signature data in the time domain, labeled them accordingly and created the 50 thousand samples dataset for DL models. All the DL models are trained and validated with a suitable number of architecture layers. By simulation, the multiclass confusion matrix, precision, recall, and F1-score are obtained in several conditions. The result shows that the stator current signature of the motor can be used to detect individual and multiple faults. Moreover, deep learning models can efficiently classify the induction motor faults based on time-domain data of the stator current signature. In deep learning (DL) models, the LSTM has shown better accuracy among all other three models. These results show that employing deep learning in fault detection and identification of induction motors can be very useful in predictive maintenance to avoid shutdown and production cycle stoppage in the industry.

Genome sequences indicate a wide variety of characteristics, which include species and sub-species type, genotype, diseases, growth indicators, yield quality, etc. To analyze and study the characteristics of the genome sequences across different species, various deep learning models have been proposed by researchers, such as Convolutional Neural Networks (CNNs), Deep Belief Networks (DBNs), Multilayer Perceptrons (MLPs), etc., which vary in terms of evaluation performance, area of application and species that are processed. Due to a wide differentiation between the algorithmic implementations, it becomes difficult for research programmers to select the best possible genome processing model for their application. In order to facilitate this selection, the paper reviews a wide variety of such models and compares their performance in terms of accuracy, area of application, computational complexity, processing delay, precision and recall. Thus, in the present review, various deep learning and machine learning models have been presented that possess different accuracies for different applications. For multiple genomic data, Repeated Incremental Pruning to Produce Error Reduction with Support Vector Machine (Ripper SVM) outputs 99.7% of accuracy, and for cancer genomic data, it exhibits 99.27% of accuracy using the CNN Bayesian method. Whereas for Covid genome analysis, Bidirectional Long Short-Term Memory with CNN (BiLSTM CNN) exhibits the highest accuracy of 99.95%. A similar analysis of precision and recall of different models has been reviewed. Finally, this paper concludes with some interesting observations related to the genomic processing models and recommends applications for their efficient use.

ObjectivesDespite showing encouraging outcomes, the precision of deep learning (DL) models using different convolutional neural networks (CNNs) for diagnosis remains under investigation. This systematic review aims to summarise the status of DL MRI models developed for assisting the diagnosis of a variety of knee abnormalities.Materials and methodsFive databases were systematically searched, employing predefined terms such as ‘Knee AND 3D AND MRI AND DL’. Selected inclusion criteria were used to screen publications by title, abstract, and full text. The synthesis of results was performed by two independent reviewers.ResultsFifty-four articles were included. The studies focused on anterior cruciate ligament injuries (n = 19, 36%), osteoarthritis (n = 9, 17%), meniscal injuries (n = 13, 24%), abnormal knee appearance (n = 11, 20%), and other (n = 2, 4%). The DL models in this review primarily used the following CNNs: ResNet (n = 11, 21%), VGG (n = 6, 11%), DenseNet (n = 4, 8%), and DarkNet (n = 3, 6%). DL models showed high-performance metrics compared to ground truth. DL models for the detection of a specific injury outperformed those by up to 4.5% for general abnormality detection.ConclusionDespite the varied study designs used among the reviewed articles, DL models showed promising outcomes in the assisted detection of selected knee pathologies by MRI. This review underscores the importance of validating these models with larger MRI datasets to close the existing gap between current DL model performance and clinical requirements.Key PointsQuestionWhat is the status of DL model availability for knee pathology detection in MRI and their clinical potential?FindingsPathology-specific DL models reported higher accuracy compared to DL models for the detection of general abnormalities of the knee. DL model performance was mainly influenced by the quantity and diversity of data available for model training.Clinical relevanceThese findings should encourage future developments to improve patient care, support personalised diagnosis and treatment, optimise costs, and advance artificial intelligence-based medical imaging practices.Graphical

Speech Emotion Recognition (SER), the extraction of emotional features with the appropriate classification from speech signals, has recently received attention for its emerging social applications. Emotional intensity (e.g., Normal, Strong) for a particular emotional expression (e.g., Sad, Angry) has a crucial influence on social activities. A person with intense sadness or anger may fall into severe disruptive action, eventually triggering a suicidal or devastating act. However, existing Deep Learning (DL)-based SER models only consider the categorization of emotion, ignoring the respective emotional intensity, despite its utmost importance. In this study, a novel scheme for Recognition of Emotion with Intensity from Speech (REIS) is developed using the DL model by integrating three speech signal transformation methods, namely Mel-frequency Cepstral Coefficient (MFCC), Short-time Fourier Transform (STFT), and Chroma STFT. The integrated 3D form of transformed features from three individual methods is fed into the DL model. Moreover, under the proposed REIS, both the single and cascaded frameworks with DL models are investigated. A DL model consists of a 3D Convolutional Neural Network (CNN), Time Distribution Flatten (TDF) layer, and Bidirectional Long Short-term Memory (Bi-LSTM) network. The 3D CNN block extracts convolved features from 3D transformed speech features. The convolved features were flattened through the TDF layer and fed into Bi-LSTM to classify emotion with intensity in a single DL framework. The 3D transformed feature is first classified into emotion categories in the cascaded DL framework using a DL model. Then, using a different DL model, the intensity level of the identified categories is determined. The proposed REIS has been evaluated on the Ryerson Audio-Visual Database of Emotional Speech and Song (RAVDESS) benchmark dataset, and the cascaded DL framework is found to be better than the single DL framework. The proposed REIS method has shown remarkable recognition accuracy, outperforming related existing methods.

Stock exchange analysis is regarded as a stochastic and demanding real-world setting in which fluctuations in stock prices are influenced by a wide range of aspects and events. In recent years, there has been a great deal of interest in social media-based data analytics for analyzing stock exchange markets. This is due to the fact that the sentiments around major global events like Brexit or COVID-19 significantly affect business decisions and investor perceptions, as well as transactional trading statistics and index values. Hence, in this research, we examined a case study from the Brexit event to assess the influence that feelings on the subject have had on the stock markets of European Union (EU) nations. Brexit has implications for Britain and other countries under the umbrella of the European Union (EU). However, a common point of debate is the EU’s contribution preferences and benefit imbalance. For this reason, the Brexit event and its impact on stock markets for major contributors and countries with minimum donations need to be evaluated accurately. As a result, to achieve accurate analysis of the stock exchanges of different EU nations from two different viewpoints, i.e., the major contributors and countries contributing least, in response to the Brexit event, we suggest an optimal deep learning and machine learning model that incorporates social media sentiment analysis regarding Brexit to perform stock market prediction. More precisely, the machine learning-based models include support vector machines (SVM) and linear regression (LR), while convolutional neural networks (CNNs) are used as a deep learning model. In addition, this method incorporates around 1.82 million tweets regarding the major contributors and countries contributing least to the EU budget. The findings show that sentiment analysis of Brexit events using a deep learning model delivers better results in comparison with machine learning models, in terms of root mean square values (RMSE). The outcomes of stock exchange analysis for the least contributing nations in relation to the Brexit event can aid them in making stock market judgments that will eventually benefit their country and improve their poor economies. Likewise, the results of stock exchange analysis for major contributing nations can assist in lowering the possibility of loss in relation to investments, as well as helping them to make effective decisions.

The increase in the number of undesired SMS termed smishing message and the data imbalance problem has generated a great demand for the development of more reliable anti‐spam filters. State of the art machine learning approaches are being employed to recognize and separate spam messages. Most recent studies target message classification by using numerous properties and features of the words but fail to consider the circumstantial features like long‐range dependencies between the words that are extremely important in identifying smishing messages. The idea is to develop an intelligent model that will distinguish between smishing messages and ham messages, by adopting a combined approach of regular expression (Regex), machine learning (ML) and deep learning (DL) models. Regex rules are generated using the dataset's spam messages for the purpose of refining the dataset. Support vector machine (SVM), Multinomial Naive Bayes and Random Forest are included under machine learning models and long short‐term memory (LSTM), bidirectional long short‐term memory (Bi‐LSTM), stacked LSTM and stacked Bi‐LSTM are included under deep learning models. The comparison between machine learning models and deep learning models is also carried out based on the performance evaluation parameters namely accuracy, precision, recall and F1 score of the models. It is observed that deep learning models perform better than machine learning models and the introduction of a regular expression to the dataset increases the efficiency of both the deep learning models and machine learning models.