Graph-Based Fusion of Imaging and Non-Imaging Data for Disease Trajectory Prediction

This study proposes a graph convolutional neural networks (GCN) architecture for fusion of radiological imaging and non-imaging tabular electronic health records (EHR) for the purpose of clinical event prediction. We focused on a cohort of hospitalized patients with positive RT-PCR test for COVID-19 and developed GCN based models to predict three dependent clinical events (discharge from hospital, admission into ICU, and mortality) using demographics, billing codes for procedures and diagnoses and chest X-rays. We hypothesized that the two-fold learning opportunity provided by the GCN is ideal for fusion of imaging information and tabular data as node and edge features, respectively. Our experiments indicate the validity of our hypothesis where GCN based predictive models outperform single modality and traditional fusion models. We compared the proposed models against two variations of imaging-based models, including DenseNet-121 architecture with learnable classification layers and Random Forest classifiers using disease severity score estimated by pre-trained convolutional neural network. GCN based model outperforms both imaging-only methods. We also validated our models on an external dataset where GCN showed valuable generalization capabilities. We noticed that edge-formation function can be adapted even after training the GCN model without limiting application scope of the model. Our models take advantage of this fact for generalization to external data. © 2023 IEEE.

COVID-19 Instance Segmentation Network Based on Semantic Information Generating Proposals

The CT images of lungs of COVID-19 patients have distinct pathological features, segmenting the lesion area accurately by the method of deep learning, which is of great significance for the diagnosis and treatment of COVID-19 patients. Instance segmentation has higher sensitivity and can output the Bounding Boxes of the lesion region, however, the traditional instance segmentation method is weak in the segmentation of small lesions, and there is still room for improvement in the segmentation accuracy. We propose a instance segmentation network which is called as Semantic R-CNN. Firstly, a semantic segmentation branch is added on the basis of Mask-RCNN, and utilizing the image processing tool Skimage in Python to label the connected domain for the result of semantic segmentation, extracting the rectangular boundaries of connected domain and using them as Proposals, which will replace the Regional Proposal Network in the instance segmentation. Secondly, the Atrous Spatial Pyramid Pooling is introduced into the Feature Pyramid Network, then improving the feature fusion method in FPN. Finally, the cascade method is introduced into the detection branch of the network to optimize the Proposals. Segmentation experiments were carried out on the pathological lesion segmentation data set of CC-CCII, the average accuracy of the semantic segmentation is 40.56mAP, and compared with the Mask-RCNN, it has improved by 9.98mAP. After fusing the results of semantic segmentation and instance segmentation, the Dice coefficient is 80.7%, the sensitivity is 85.8%, and compared with the Inf-Net, it has increased by 1.6% and 8.06% respectively. The proposed network has improved the segmentation accuracy and reduced the false-negatives. © 2022 ACM.

An Automatic CNN-based Face Mask Detection Algorithm Tested During the COVID-19 Pandemics

The ongoing COVID-19 pandemic has highlighted the importance of wearing face masks as a preventive measure to reduce the spread of the virus. In medical settings, such as hospitals and clinics, healthcare professionals and patients are required to wear surgical masks for infection control. However, the use of masks can hinder facial recognition technology, which is commonly used for identity verification and security purposes. In this paper, we propose a convolutional neural network (CNN) based approach to detect faces covered by surgical masks in medical settings. We evaluated the proposed CNN model on a test set comprising of masked and unmasked faces. The results showed that our model achieved an accuracy of over 96% in detecting masked faces. Furthermore, our model demonstrated robustness to different mask types and fit variations commonly encountered in medical settings. Our approaches reaches state of the art results in terms of accuracy and generalization. © 2022 Copyright for this paper by its authors.

Deep Convolutional Neural Network Based Automatic COVID-19 Detection from Chest X-ray Images

To minimize the rate of death from COVID-19 and stop the disease from spreading early detection is vital. The normal RT-PCR tests for COVID-19 detection take a long time to complete. In contrast to this test, Covid-19 can be quickly detected using various machine-learning technologies. Previous studies only had access to smaller datasets, as COVID-19 data was not readily available back then. Since COVID-19 is a dangerous virus, the model needs to be robust and trustworthy, and the model must be trained on a large and diverse dataset. To overcome that problem, this study combines six publicly available Chest X-ray datasets to produce a larger and more diverse balanced dataset with a total of 68,424 images. In this study, we develop a CNN model that primarily entails two steps: (a) feature extraction and (b) classification, which are used to identify COVID-19 positive cases from X-ray images. The accuracy of this proposed model is 97.58%, which is higher than most state-of-the-art models. © 2022 IEEE.

BEsafe - Validating URLs and Domains with the aid of Indicator of Compromise

The advancement of information technology has stimulated the conversion of physical interactions to online activities, especially during the Covid-19 pandemic. Thus, users' awareness and cyber hygiene need to be emphasized when they are involved in the cyber world. A browser extension named 'BEsafe' is developed to validate the websites and promote a safe browsing environment. It prevents users from falling prey to network-based attacks and raises their security awareness. To ensure users' privacy, the permissions needed for BEsafe are listed on the permission tab. Moreover, BEsafe will not be working on Incognito mode by default to promise that the private mode leaves no tracks. However, the user can still enable the extension to be functioning on Incognito mode by navigating to the Extension Details and turning on the relevant toggle. © 2023 IEEE.

GAN-based Algorithm for Efficient Image Inpainting

Global pandemic due to the spread of COVID-19 has post challenges in a new dimension on facial recognition, where people start to wear masks. Under such condition, the authors consider utilizing machine learning in image inpainting to tackle the problem, by complete the possible face that is originally covered in mask. In particular, autoencoder has great potential on retaining important, general features of the image as well as the generative power of the generative adversarial network (GAN). The authors implement a combination of the two models, context encoders and explain how it combines the power of the two models and train the model with 50,000 images of influencers faces and yields a solid result that still contains space for improvements. Furthermore, the authors discuss some shortcomings with the model, their possible improvements, as well as some area of study for future investigation for applicative perspective, as well as directions to further enhance and refine the model. © 2023 SPIE.

MLCNN-COV: A multilabel convolutional neural network-based framework to identify negative COVID medicine responses from the chemical three-dimensional conformer

To treat the novel COronaVIrus Disease (COVID), comparatively fewer medicines have been approved. Due to the global pandemic status of COVID, several medicines are being developed to treat patients. The modern COVID medicines development process has various challenges, including predicting and detecting hazardous COVID medicine responses. Moreover, correctly predicting harmful COVID medicine reactions is essential for health safety. Significant developments in computational models in medicine development can make it possible to identify adverse COVID medicine reactions. Since the beginning of the COVID pandemic, there has been significant demand for developing COVID medicines. Therefore, this paper presents the transfer-learning methodology and a multilabel convolutional neural network for COVID (MLCNN-COV) medicines development model to identify negative responses of COVID medicines. For analysis, a framework is proposed with five multilabel transfer-learning models, namely, MobileNetv2, ResNet50, VGG19, DenseNet201, and Inceptionv3, and an MLCNN-COV model is designed with an image augmentation (IA) technique and validated through experiments on the image of three-dimensional chemical conformer of 17 number of COVID medicines. The RGB color channel is utilized to represent the feature of the image, and image features are extracted by employing the Convolution2D and MaxPooling2D layer. The findings of the current MLCNN-COV are promising, and it can identify individual adverse reactions of medicines, with the accuracy ranging from 88.24% to 100%, which outperformed the transfer-learning model's performance. It shows that three-dimensional conformers adequately identify negative COVID medicine responses. 1225-6463/$ © 2023 ETRI.

Accuration of Classification of Covid with Convolutional Neural Network-Based Image Chest X-ray with Variations in Image Size and Batch Size

The use of convolutional neural networks in Covid classification has a positive impact on the speed of justification and can provide high accuracy. But on the one hand, the many parameters on CNN will also have an impact on the resulting accuracy. CNN requires time and a heavy level of computation. Setting the right parameters will provide high accuracy. This study examines the performance of CNN with variations in image size and minibatch. Parameter settings used are max epoch values of 100, minibatch variations of 32, 64, and 128, and learning rate of 0.1 with image size inputs of 50,100, and 150 variations on the level of accuracy. The dataset consists of training data and test data, 200 images, which are divided into two categories of normal and abnormal images (Covid). The results showed an accuracy with the use of minibatch 128 with the highest level of accuracy at image size 150 × 150 on test data of 99,08%. The size of the input matrix does not always have an impact on increasing the level of accuracy, especially on the minibatch 32. The parameter setting on CNN was dependent on the CNN architecture, the dataset used, and the size of the dataset. One can imply that optimization parameter in CNN can approve good accuration. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

COVID-19 Contact-Tracing Networks Datasets

In this research paper, COVID-19 tracing data are utilized to form two dataset networks, one is based on the virus transition between the world countries, as the dataset consists of 36 countries and 75 relationships between them. Whereas the other dataset is an attributed network based on the virus transition among the contact tracing in the Kingdom of Bahrain. This type of networks that is concerned in tracking a disease or virus was not formed based on COVID-19 virus transmission. © 2023 IEEE.

Croup disease classification using VGG19 and ResNet50 transfer learning method

Detection of any disease in the early stage can save a life. There are many medical imaging modalities like MRI, FMRI, ultrasound, CT, and X-ray used in the detection of disease. In the last decades, neural network-based methods are effective in detecting and classifying the disease based on abnormalities present in the medical images. Acute laryngotracheobronchitis (croup) is one of the common diseases seen in children among the 0.5-3 years age group which infects the respiratory system that can cause the larynx, trachea, and bronchi. Prior detection can lower the risk of spreading and can be treated accurately by a pediatrician. Commonly this infection can be diagnosed though physical examination. But due to the similarity of Covid-19 symptoms urges the physicians to get accurate detection of this disease using X-ray and CT images of the infant's chest and throat. The proposed work aims to develop a croup diagnose system (CDS) which identify the Croup infection through post anterior (PA) view of pediatric X-ray using deep learning algorithm. We used the well-known transfer learning algorithm VGG19 and ResNet50. Data augmentation being adapted for reducing the overfitting and to improve the quantity of image samples. We show that the proposed transfer learning based CDS method can be used to classify the X-ray images into two classes namely, croup and normal. The experiment results confirm that VGG19 performs better than ResNet50 with promising classification accuracy (90.91%.). The results show that the proposed CDS models can be used for more pediatric medical image classification problem. © 2024 World Scientific Publishing Company.

Target prediction, computational identification, and network-based pharmacology of most potential phytoconstituent in medicinal leaves of Justicia adhatoda against SARS-CoV-2.

The current global epidemic of the novel coronavirus (SARS-CoV-2) has been labeled a global public health emergency since it is causing substantial morbidity and mortality on daily basis. We need to identify an effective medication against SARS-CoV-2 because of its fast dissemination and re-emergence. This research is being carried out as part of a larger strategy to identify the most promising therapeutic targets using protein-protein interactions analysis. Mpro has been identified as one of the most important therapeutic targets. In this study, we did in-silico investigations to identify the target and further molecular docking, ADME, and toxicity prediction were done to assess the potential phyto-active antiviral compounds from Justicia adhatoda as powerful inhibitors of the Mpro of SARS-COV-2. We also investigated the capacity of these molecules to create stable interactions with the Mpro using 100 ns molecular dynamics simulation. The highest scoring compounds (taraxerol, friedelanol, anisotine, and adhatodine) were also found to exhibit excellent solubility and pharmacodynamic characteristics. We employed MMPBSA simulations to assess the stability of docked molecules in the Mpro binding site, revealing that the above compounds form the most stable complex with the Mpro. Network-based Pharmacology suggested that the selected compounds have various modes of action against SARS-CoV-2 that include immunoreaction enrichment, inflammatory reaction suppression, and more. These findings point to a promising class of drugs that should be investigated further in biochemical and cell-based studies to see their effectiveness against nCOVID-19.Communicated by Ramaswamy H. Sarma.

Smart Methods to Deal with COVID-19 at University-Level Institutions Using Social Network Analysis Techniques

The current global health crisis is a consequence of the pandemic caused by COVID-19. It has impacted the lives of people from all factions of society. The re-emergence of new variants is threatening the world, which urges the development of new methods to prevent rapid spread. Places with more extensive social dealings, such as offices, organizations, and educational institutes, have a greater tendency to escalate the viral spread. This research focuses on developing a strategy to find out the key transmitters of the virus, particularly at educational institutes. The reason for considering educational institutions is the severity of the educational needs and the high risk of rapid spread. Educational institutions offer an environment where students come from different regions and communicate with each other at close distances. To slow down the virus's spread rate, a method is proposed in this paper that differs from vaccinating the entire population or complete lockdown. In the present research, we identified a few key spreaders, which can be isolated and can slow down the transmission rate of the contagion. The present study creates a student communication network, and virus transmission is modeled over the predicted network. Using student-to-student communication data, three distinct networks are generated to analyze the roles of nodes responsible for the spread of this contagion. Intra-class and inter-class networks are generated, and the contagion spread was observed on them. Using social network strategies, we can decrease the maximum number of infections from 200 to 70 individuals, with contagion lasting in the network for 60 days.

Speech Recognition and Neural Networks based Talking Health Care Bot (THCB): Medibot

The COVID-19 pandemic has affected healthcare in several ways. Some patients were unable to make it to appointments due to curfews, transportation restrictions, and stay-at-home directives, while less urgent procedures were postponed or cancelled. Others steered clear of hospitals out of fear of contracting an infection. With the use of a conversational artificial intelligence-based program, the Talking Health Care Bot (THCB) could be useful during the pandemic by allowing patients to receive supportive care without physically visiting a hospital. Therefore, the THCB will drastically and quickly change in-person care to patient consultation through the internet. To give patients free primary healthcare and to narrow the supply-demand gap for human healthcare professionals, this work created a conversational bot based on artificial intelligence and machine learning. The study proposes a revolutionary computer program that serves as a patient's personal virtual doctor. The program was carefully created and thoroughly trained to communicate with patients as if they were real people. Based on a serverless architecture, this application predicts the disease based on the symptoms of the patients. A Talking Healthcare chatbot confronts several challenges, but the user's accent is by far the most challenging. This study has then evaluated the proposed model by using one hundred different voices and symptoms, achieving an accuracy rate of 77%. © 2023 IEEE.

Network based Learning Platform Application Model for Enhancement of Realtime Working Systems

Learning platforms have become an integral part of the education system these days. Especially after the COVID era, education has become more allied with self-paced and remote learning and learning platforms have made it boom exponentially. Improper construction and implementation of such platforms can lead to huge risks for the users and the company. Data security is not taken much care of while building such platforms;instead, concentration is given to fancy front-end pages and attractive interfaces. This may not be good at all times. Data is one of the most powerful resources and can have a very big impact if misused. This paper proposes a networking-based approach to implementing such platform systems in a safe and organized way. Implementation using networking concepts gives a better hand in managing permissions, access rights, and security in all data-related transfers and communications. In terms of online gaming and real-time video communication, User Datagram Protocol (UDP) is often used because it is faster than Transmission Control Protocol (TCP) and is well-suited for real-time applications that cannot tolerate delays. UDP is a connectionless protocol, which means it doesn't retransmit lost data and therefore has lower overhead, making it a good choice for real-time applications like video conferencing and online gaming. Examples of such applications include Skype, Google Meet, Zoom, and Facetime. Based on these existing applications, this work introduces UDP in the field of Learning Platform Applications and builds a model on top of which real-time applications can be constructed. The proposed system makes use of UDP for all its requests, responses, and file transfers. The protocol itself is not very reliable, but the addition of provisions for acknowledgements in all requests and responses makes this system overcome transfer uncertainty. Implementation using networking concepts improves the speed, security, privacy, and customization abilities of the proposed system. © 2023 IEEE.

Dynamic spillovers between clean energy and non-ferrous metals markets in China: A network-based analysis during the COVID-19 pandemic

Due to the close production link between clean energy and non-ferrous metals, their price and market dynamics can easily affect one another through production costs. Furthermore, with the increased financialization of clean energy and non-ferrous metals markets, investment risk can easily spread between them. Therefore, this paper intends to explore the risk contagion between the two markets through the spillover index model and the minimum spanning tree (MST) method. Employing the data collected in China, this paper quantifies the magnitude of risk transfer by the volatility spillovers of eight clean energy stock markets as identified in The Energy Conservation and Environmental Protection Clean Industry Statistical Classification 2021 and the eight corresponding non-ferrous metals futures markets, while fully considering the heterogeneity between sub-markets. First, we find that risk is mainly transmitted from clean energy to non-ferrous metals. Second, this paper identifies not only the most influential market but also the shortest path of risk contagion based on the MST topology analysis. Last, the empirical results show that the COVID-19 has increased the scale of risk transmission between the two markets and their connectivity. During the COVID-19 period, the shortest path between the two markets shifted from "hydropower–gold” to "smart grid–zinc”, and the systematically influential markets correspondingly become smart grid and zinc. The results obtained in this paper might have practical implications for policymakers seeking to achieve effective risk management, which could also facilitate investors for diversification benefits. © 2023 Elsevier Ltd

Comparitive Analysis on Neural Networks based on their performance in Pneumonia Detection

Pneumonia has been a concerning issue worldwide. This infectious disease has a higher mortality rate than Covid-19. More than two million individuals lost their lives in 2019 out of which almost 600,000 were infants less than 5 years of age. Globally, identification of the disease is done manually by radiologists, but this method is highly unreliable as its accuracy is not sufficiently good. With the evolution of computational resources, especially the computing power of GPUs, it has become possible to train very deep CNNs. This study involves a comparative analysis of neural networks for pneumonia recognition. The goal is to do binary image classification for pneumonia recognition on each of the three models, namely, a Sequential model using TensorFlow (built from scratch), ResNet50 and InceptionV3 and comparing their efficiency, to discover which model suits best for smaller datasets and which suits best for larger datasets. Dataset consists of 5856 anterior and posterior Chest X-Ray images labeled as either Normal or Pneumonic. © 2023 IEEE.

Forecasting COVID-19 Cases Using n-SARS-CoV-2 Variants

This work proposes a novel Deep Learning-based model to forecast the total number of confirmed COVID-19 cases in four of the worst-hit states of India. Along with statewide restrictions and public holidays, a novel parameter is introduced for training the proposed model, which considers the Alpha, Beta, Delta, and Omicron variants and the degree of their prevalence in each of the four states. Recurrent Neural Network-based Long-Short Term Memory is applied to the custom dataset, with the lowest Mean Absolute Percentage Error being 0.77% for the state of Maharashtra. SHapley Additive exPlanations values are used to examine the significance of the various parameters. The proposed model can be applied to other countries and can include newer variants of the novel coronavirus discovered in the future. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Crowd Density Estimation Using Neural Network for COVID-19 and Future Pandemics

The population's vulnerability is exacerbated by the lack of effective treatment drugs and immunity to COVID-19. The only viable strategy for combating this pandemic is social separation. In order to automate the task of monitoring social separation using surveillance footage, this study presents a neural network-based crowd density estimation for COVID-19 and future pandemics. The suggested framework employs the object identification model to distinguish persons in the scene, as well as the deep sort technique to track recognized people with issued IDs. The obtained results of the proposed work are compared in terms of loss values defined by object classification and localization, frames per second (FPS), and mean average precision (mAP). The proposed method yields good results against faster region-based convolutional neural network (RCNN) and single-shot detector (SSD). © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023.

Early Detection of Covid Using Spectral Analysis of Cough and Deep Convolutional Neural Network

Now-a-days, there are numerous techniques and ICT tools for the detection of Covid-19. But, these techniques are working with the help;of culminated or peak of symptoms. However, there is a demanding need for the early detection of Covid with self-reported symptoms or even without any symptoms, which makes it easier for further diagnosis or treatment. This research paper proposes a novel approach for the early detection of Covid with the spectral analysis of Cough sound using discrete wavelet transform (DWT), followed by deep convolution neural network (DCNN) based classification. The proposed method with the cough spectral analysis and Deep Learning based algorithm returns the covid infection probability. The empirical results show that the proposed method of covid detection using cough spectral analysis using DWT and deep learning achieves better accuracy, while compared to the conventional methods. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

A Lightweight Segmentation Network Based on Weak Supervision for COVID-19 Detection

The Coronavirus Disease 2019 (COVID-19) outbreak in late 2019 threatens global health security. Computed tomography (CT) can provide richer information for the diagnosis and treatment of COVID-19. Unfortunately, labeling of COVID-19 lesion chest CT images is an expensive affair. We solved the challenge of chest CT labeling by simply marking point annotations to the lesion areas, i.e., by marking individual pixels for each lesion area in the chest CT scan. It takes only a few seconds to complete the labeling using this labeling strategy. We also designed a lightweight segmentation model with approximately 10% of the number of model parameters of the conventional model. So, the proposed model segmented the lesions of a single image in only 0.05 s. In order to obtain the shape and size of lesions from point labels, the convex-hull based segmentation (CHS) loss function is proposed in this paper, which enables the model to obtain an approximate fully supervised performance on point labels. The experiments were compared with the current state-of-the-art (SOTA) point label segmentation methods on the COVID-19-CT-Seg dataset, and our model showed a large improvement: IoU improved by 28.85%, DSC improved by 28.91%, Sens improved by 13.75%, Spes improved by 1.18%, and MAE decreased by 1.10%. Experiments on the dataset show that the proposed model combines the advantages of lightweight and weak supervision, resulting in more accurate COVID-19 lesion segmentation results while having only a 10% performance difference with the fully supervised approach. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.