Classification of deep features obtained from mpox images using Xception by selecting them with PCA and mRMR and using SVM

BackgroundChest CT is used for the assessment of the severity of patients infected with novel coronavirus 2019 (COVID-19). We collected chest CT scans of 202 patients diagnosed with the COVID-19, and try to develop a rapid, accurate and automatic tool for severity screening follow-up therapeutic treatment.MethodsA total of 729 2D axial plan slices with 246 severe cases and 483 non-severe cases were employed in this study. By taking the advantages of the pre-trained deep neural network, four pre-trained off-the-shelf deep models (Inception-V3, ResNet-50, ResNet-101, DenseNet-201) were exploited to extract the features from these CT scans. These features are then fed to multiple classifiers (linear discriminant, linear SVM, cubic SVM, KNN and Adaboost decision tree) to identify the severe and non-severe COVID-19 cases. Three validation strategies (holdout validation, tenfold cross-validation and leave-one-out) are employed to validate the feasibility of proposed pipelines.Results and conclusionThe experimental results demonstrate that classification of the features from pre-trained deep models shows the promising application in COVID-19 severity screening, whereas the DenseNet-201 with cubic SVM model achieved the best performance. Specifically, it achieved the highest severity classification accuracy of 95.20% and 95.34% for tenfold cross-validation and leave-one-out, respectively. The established pipeline was able to achieve a rapid and accurate identification of the severity of COVID-19. This may assist the physicians to make more efficient and reliable decisions.

Gesture Recognition pertains to recognizing meaningful expressions of motion by a human, involving the hands, arms, face, head, and/or body. It is of utmost importance in designing an intelligent and efficient human–computer interface. The applications of gesture recognition are manifold, ranging from sign language through medical rehabilitation, monitoring patients or elder people, surveillance systems, sports gesture analysis, human behaviour analysis etc., to virtual reality. In recent years, there has been increased interest in video summarization and automatic sports highlights generation in the game of Cricket. In Cricket, the Umpire has the authority to make important decisions about events on the field. The Umpire signals important events using unique hand on signals and gestures. The primary intention of our work is to design and develop a new robust method for Umpire Action and Non-Action Gesture Identification and Recognition based on the Umpire Segmentation and the proposed Histogram Oriented Gradient (HOG) feature Extraction oriented Non-Linear Support Vector Machine (NL-SVM) classification of Deep Features. Primarily the 80% of Umpire action and non-action images in a cricket match, about 1, 93, 000 frames, the Histogram of Oriented Gradient Deep Features are calculated and trained the system having six gestures of Umpire pose. The proposed HOG Feature Extraction oriented Non-Linear Support Vector Machine classification method achieves the maximal accuracy of 97.95%, the maximal sensitivity of 98.87%, the maximal specificity of 98.89% and maximal Precision of 97.02% which indicates its superiority.

Natural language processing (NLP) models based on deep neural networks (DNNs) are vulnerable to backdoor attacks. Existing backdoor defense methods have limited effectiveness and coverage scenarios. We propose a textual backdoor defense method based on deep feature classification. The method includes deep feature extraction and classifier construction. The method exploits the distinguishability of deep features of poisoned data and benign data. Backdoor defense is implemented in both offline and online scenarios. We conducted defense experiments on two datasets and two models for a variety of backdoor attacks. The experimental results demonstrate the effectiveness of this defense approach and outperform the baseline defense method.

ABSTRACTThis letter proposes a synthetic aperture radar (SAR) target recognition method via joint classification of deep features fused by multi-canonical correlation analysis (MCCA). A convolutional neural network (CNN) is designed for feature learning from original SAR images. For the multiple feature maps from different convolution layers, they are fused based on the MCCA to maintain the relevance while eliminating the redundancy. Afterwards, the joint sparse representation (JSR) is employed to jointly represent the fused deep feature vectors from different convolution layers under the constraint of their inner correlations. Based on the reconstruction errors from JSR, the target label can be classified. The proposed method can make full use of the multi-level deep features by using the correlations among the same layer and between different layers. Experiments are investigated on the Moving and Stationary Target Acquisition and Recognition (MSTAR) data set and the results confirm the performance of the proposed method.

Transfer learning is commonly used in deep learning for medical imaging to alleviate the problem of limited available data. In this work, we studied the risk of feature leakage and its dependence on sample size when using pretrained deep convolutional neural network (DCNN) as feature extractor for classification breast masses in mammography. Feature leakage occurs when the training set is used for feature selection and classifier modeling while the cost function is guided by the validation performance or informed by the test performance. The high-dimensional feature space extracted from pretrained DCNN suffers from the curse of dimensionality; feature subsets that can provide excessively optimistic performance can be found for the validation set or test set if the latter is allowed for unlimited reuse during algorithm development. We designed a simulation study to examine feature leakage when using DCNN as feature extractor for mass classification in mammography. Four thousand five hundred and seventy-seven unique mass lesions were partitioned by patient into three sets: 3222 for training, 508 for validation, and 847 for independent testing. Three pretrained DCNNs, AlexNet, GoogLeNet, and VGG16, were first compared using a training set in fourfold cross validation and one was selected as the feature extractor. To assess generalization errors, the independent test set was sequestered as truly unseen cases. A training set of a range of sizes from 10% to 75% was simulated by random drawing from the available training set in addition to 100% of the training set. Three commonly used feature classifiers, the linear discriminant, the support vector machine, and the random forest were evaluated. A sequential feature selection method was used to find feature subsets that could achieve high classification performance in terms of the area under the receiver operating characteristic curve (AUC) in the validation set. The extent of feature leakage and the impact of training set size were analyzed by comparison to the performance in the unseen test set. All three classifiers showed large generalization error between the validation set and the independent sequestered test set at all sample sizes. The generalization error decreased as the sample size increased. At 100% of the sample size, one classifier achieved an AUC as high as 0.91 on the validation set while the corresponding performance on the unseen test set only reached an AUC of 0.72. Our results demonstrate that large generalization errors can occur in AI tools due to feature leakage. Without evaluation on unseen test cases, optimistically biased performance may be reported inadvertently, and can lead to unrealistic expectations and reduce confidence for clinical implementation.

Lung disease is among the leading diseases that cause mortality worldwide. Most cases of lung diseases are detected when the disease is in the advanced stages. Therefore the development of systems and methods that enable faster and early diagnosis will play a vital role in the world today. Computer Aided Diagnosis (CADx) systems play such a role and are currently being expanded. This study explores the potential of using deep learning features from pre-trained deep learning architectures to provide rich and robust features. These features were compared to the conventionally used Gray-level Co-occurrence Matrix (GLCM). Deep features produced the highest accuracy of 100% as compared to 93.52% produced by using GLCM features. This study also compared the classification of deep features with five different classifiers and Support Vector Machine (SVM) showed the highest result. This high accuracy was also reproduced with Linear Discriminant Analysis (LDA) and Regression classifiers. Principal Component Analysis (PCA) was also done to evaluate the usage of reduced number of features and its effect on the classification performance. Using deep features produced 4096 features and a classification accuracy of 100%. When PCA is introduced, only 79 features were used however the accuracy produced was the same. Thus, there is promising use of deep features together with PCA to reduce the number of features in the classification of diseased lungs.

This article has been retracted by IOP Publishing following an allegation that this article may contain tortured phrases [1], masking overlap of other work. [2-4]. IOP Publishing has investigated in line with the COPE guidelines and agree this article should be retracted. IOP Publishing wishes to credit the Problematic Paper Screener for bringing the issue to our attention. The authors have not confirmed whether they agree or disagree to this retraction. [1] Cabanac G, Labbe C, Magazinov A, 2021, Tortured phrases: A dubious writing style emerging in science. Evidence of critical issues affecting established journals, arXiv:2107.06751v1 [2] R Hari and M Wilscy, 2014, Event detection in cricket videos using intensity projection profile of Umpire gestures 2014 Annual IEEE India Conference (INDICON) [3] M.H Kolekar and s Sengupta, 2010, Semantic concept mining in cricket videos for automated highlight generation Multimed Tools Appl 47 545–579 [4] S Nandyal and S.L Kattimani, 2021, An Efficient Umpire Key Frame Segmentation in Cricket Video using HOG and SVM 2021 6th International Conference for Convergence in Technology (I2CT) Retraction published: 10 February 2023

Automatic detection of left ventricle myocardium is essential to subsequent cardiac image registration and tissue segmentation. However, it is considered challenging mainly because of the complex and varying shape of the myocardium and surrounding tissues across slices and phases. In this study, a hybrid model is proposed to detect myocardium in cardiac magnetic resonance (MR) images combining region proposal and deep feature classification and regression. The model firstly generates candidate regions using new structural similarity-enhanced supervoxel over-segmentation plus hierarchical clustering. Then it adopts a deep stacked sparse autoencoder (SSAE) network to learn the discriminative deep feature to represent the regions. Finally, the features are fed to train a novel nonlinear within-class neighborhood preserved soft margin support vector (C-SVC) classifier and multiple-output support vector (-SVR) regressor for refining the location of myocardium. To improve the stability and generalization, the model also takes hard negative sample mining strategy to fine-tune the SSAE and the classifier. The proposed model with impacts of different components were extensively evaluated and compared to related methods on public cardiac data set. Experimental results verified the effectiveness of proposed integrated components, and demonstrated that it was robust in myocardium localization and outperformed the state-of-the-art methods in terms of typical metrics. This study would be beneficial in some cardiac image processing such as region-of-interest cropping and left ventricle volume measurement.

To develop an automatic deep feature classification (DFC) method for distinguishing benign angiomyolipoma without visible fat (AMLwvf) from malignant clear cell renal cell carcinoma (ccRCC) from abdominal contrast-enhanced computer tomography (CE CT) images. A dataset including 80 abdominal CT images of 39 AMLwvf and 41 ccRCC patients was used. We proposed a DFC method for differentiating the small renal masses (SRM) into AMLwvf and ccRCC using the combination of hand-crafted and deep features, and machine learning classifiers. First, 71-dimensional hand-crafted features (HCF) of texture and shape were extracted from the SRM contours. Second, 1000-4000-dimensional deep features (DF) were extracted from the ImageNet pretrained deep learning model with the SRM image patches. In DF extraction, we proposed the texture image patches (TIP) to emphasize the texture information inside the mass in DFs and reduce the mass size variability. Finally, the two features were concatenated and the random forest (RF) classifier was trained on these concatenated features to classify the types of SRMs. The proposed method was tested on our dataset using leave-one-out cross-validation and evaluated using accuracy, sensitivity, specificity, positive predictive values (PPV), negative predictive values (NPV), and area under receiver operating characteristics curve (AUC). In experiments, the combinations of four deep learning models, AlexNet, VGGNet, GoogleNet, and ResNet, and four input image patches, including original, masked, mass-size, and texture image patches, were compared and analyzed. In qualitative evaluation, we observed the change in feature distributions between the proposed and comparative methods using tSNE method. In quantitative evaluation, we evaluated and compared the classification results, and observed that (a) the proposed HCF+DF outperformed HCF-only and DF-only, (b) AlexNet showed generally the best performances among the CNN models, and (c) the proposed TIPs not only achieved the competitive performances among the input patches, but also steady performance regardless of CNN models. As a result, the proposed method achieved the accuracy of 76.6±1.4% for the proposed HCF+DF with AlexNet and TIPs, which improved the accuracy by 6.6%p and 8.3%p compared to HCF-only and DF-only, respectively. The proposed shape features and TIPs improved the HCFs and DFs, respectively, and the feature concatenation further enhanced the quality of features for differentiating AMLwvf from ccRCC in abdominal CE CT images.

Coronavirus19 is caused due to infection in the respiratory system. It is the type of RNA virus that might infect animal and human species. In the severe stage, it causes pneumonia in human beings. In this research, hand‐crafted and deep microscopic features are used to classify lung infection. The proposed work consists of two phases; in phase I, infected lung region is segmented using proposed U‐Net deep learning model. The hand‐crafted features are extracted such as histogram orientation gradient (HOG), noise to the harmonic ratio (NHr), and segmentation based fractal texture analysis (SFTA) from the segmented image, and optimum features are selected from each feature vector using entropy. In phase II, local binary patterns (LBPs), speeded up robust feature (Surf), and deep learning features are extracted using a pretrained network such as inceptionv3, ResNet101 from the input CT images, and select optimum features based on entropy. Finally, the optimum selected features using entropy are fused in two ways, (i) The hand‐crafted features (HOG, NHr, SFTA, LBP, SURF) are horizontally concatenated/fused (ii) The hand‐crafted features (HOG, NHr, SFTA, LBP, SURF) are combined/fused with deep features. The fused optimum features vector is passed to the ensemble models (Boosted tree, bagged tree, and RUSBoosted tree) in two ways for the COVID19 classification, (i) classification using fused hand‐crafted features (ii) classification using fusion of hand‐crafted features and deep features. The proposed methodology is tested /evaluated on three benchmark datasets. Two datasets employed for experiments and results show that hand‐crafted & deep microscopic feature's fusion provide better results compared to only hand‐crafted fused features.

Microgrids are an essential element of smart grids, which contain distributed renewable energy sources (RESs), energy storage devices, and load control strategies. Models built based on machine learning (ML) and deep learning (DL) offer hope for anticipating consumer demands and energy production from RESs. This study suggests an innovative approach for energy analysis based on the feature extraction and classification of microgrid photovoltaic cell data using deep learning algorithms. The energy optimization of a microgrid was carried out using a photovoltaic energy system with distributed power generation. The data analysis has been carried out for feature analysis and classification using a Gaussian radial Boltzmann with Markov encoder model. Based on microgrid energy optimization and data analysis, an experimental analysis of power analysis, energy efficiency, quality of service (QoS), accuracy, precision, and recall has been conducted. The proposed technique attained power analysis of 88%, energy efficiency of 95%, QoS of 77%, accuracy of 93%, precision of 85%, and recall of 77%.

Convolutional Neural Networks (ConvNets) are increasingly being used for medical image diagnostic applications. In this paper, we compare two transfer learning approaches - Deep Feature classification and Fine-tuning ConvNets for Diagnosing Breast Cancer malignancy. BreaKHis dataset is used to benchmark our results with ResNet-50, InceptionV2 and DenseNet-169 pre-trained models. Deep feature classification accuracy ranges from 81% to 95% using Logistic Regression, LightGBM and Random Forest classifiers. Fine-tuned DenseNet-169 model accuracy outperformed all other classification models with 99.25 ± 0.4%.

Deep time-frequency features and semi-supervised dimension reduction for subject-independent emotion recognition from multi-channel EEG signals

Magnetic resonance imaging (MRI) is increasingly used in the diagnosis of Alzheimer's disease (AD) in order to identify abnormalities in the brain. Indeed, cortical atrophy, a powerful biomarker for AD, can be detected using structural MRI (sMRI), but it cannot detect impairment in the integrity of the white matter (WM) preceding cortical atrophy. The early detection of these changes is made possible by the novel MRI modality known as diffusion tensor imaging (DTI). In this study, we integrate DTI and sMRI as complementary imaging modalities for the early detection of AD in order to create an effective computer‐assisted diagnosis tool. The fused Bag‐of‐Features (BoF) with Speeded‐Up Robust Features (SURF) and modified AlexNet convolutional neural network (CNN) are utilized to extract local and deep features. This is applied to DTI scalar metrics (fractional anisotropy and diffusivity metric) and segmented gray matter images from T1‐weighted MRI images. Then, the classification of local unimodal and deep multimodal features is first performed using support vector machine (SVM) classifiers. Then, the majority voting technique is adopted to predict the final decision from the ensemble SVMs. The study is directed toward the classification of AD versus mild cognitive impairment (MCI) versus cognitively normal (CN) subjects. Our proposed method achieved an accuracy of 98.42% and demonstrated the robustness of multimodality imaging fusion.

The immune system is one of the most critical systems in humans that resists all diseases and protects the body from viruses, bacteria, etc. White blood cells (WBCs) play an essential role in the immune system. To diagnose blood diseases, doctors analyze blood samples to characterize the features of WBCs. The characteristics of WBCs are determined based on the chromatic, geometric, and textural characteristics of the WBC nucleus. Manual diagnosis is subject to many errors and differing opinions of experts and takes a long time; however, artificial intelligence techniques can help to solve all these challenges. Determining the type of WBC using automatic diagnosis helps hematologists to identify different types of blood diseases. This work aims to overcome manual diagnosis by developing automated systems for classifying microscopic blood sample datasets for the early detection of diseases in WBCs. Several proposed systems were used: first, neural network algorithms, such as artificial neural networks (ANNs) and feed-forward neural networks (FFNNs), were applied to diagnose the dataset based on the features extracted using the hybrid method between two algorithms, the local binary pattern (LBP) and gray-level co-occurrence matrix (GLCM). All algorithms attained superior accuracy for WBC diagnosis. Second, the pre-trained convolutional neural network (CNN) models AlexNet, ResNet-50, GoogLeNet, and ResNet-18 were applied for the early detection of WBC diseases. All models attained exceptional results in the early detection of WBC diseases. Third, the hybrid technique was applied, consisting of a pair of blocks: the CNN models block for extracting deep features and the SVM algorithm block for the classification of deep features with superior accuracy and efficiency. These hybrid techniques are named AlexNet with SVM, ResNet-50 with SVM, GoogLeNet with SVM, and ResNet-18 with SVM. All techniques achieved promising results when diagnosing the dataset for the early detection of WBC diseases. The ResNet-50 model achieved an accuracy of 99.3%, a precision of 99.5%, a sensitivity of 99.25%, a specificity of 99.75%, and an AUC of 99.99%.