Hybrid Models for Breast Cancer Detection via Transfer Learning Technique

Currently,breast cancer has been amajor cause of deaths in women worldwide and the World Health Organization(WHO)has confirmed this.The severity of this disease can be minimized to the large extend,if it is diagnosed properly at an early stage of the disease.Therefore,the proper treatment of a patient having cancer can be processed in better way,if it can be diagnosed properly as early as possible using the better algorithms.Moreover,it has been currently observed that the deep neural networks have delivered remarkable performance for detecting cancer in histopathological images of breast tissues.To address the above said issues,this paper presents a hybrid model using the transfer learning to study the histopathological images,which help in detection and rectification of the disease at a low cost.Extensive dataset experiments were carried out to validate the suggested hybrid model in this paper.The experimental results show that the proposed model outperformed the baseline methods,with F-scores of 0.81 for DenseNet+Logistic Regression hybrid model,(F-score:0.73)for Visual Geometry Group(VGG)+Logistic Regression hybrid model,(F-score:0.74)for VGG+Random Forest,(F-score:0.79)for DenseNet+Random Forest,and(F-score:0.79)for VGG+Densenet+Logistic Regression hybrid model on the dataset of histopathological images.

Two-Stream Deep Learning Architecture-Based Human Action Recognition

Human action recognition(HAR)based on Artificial intelligence reasoning is the most important research area in computer vision.Big breakthroughs in this field have been observed in the last few years;additionally,the interest in research in this field is evolving,such as understanding of actions and scenes,studying human joints,and human posture recognition.Many HAR techniques are introduced in the literature.Nonetheless,the challenge of redundant and irrelevant features reduces recognition accuracy.They also faced a few other challenges,such as differing perspectives,environmental conditions,and temporal variations,among others.In this work,a deep learning and improved whale optimization algorithm based framework is proposed for HAR.The proposed framework consists of a few core stages i.e.,frames initial preprocessing,fine-tuned pre-trained deep learning models through transfer learning(TL),features fusion using modified serial based approach,and improved whale optimization based best features selection for final classification.Two pre-trained deep learning models such as InceptionV3 and Resnet101 are fine-tuned and TL is employed to train on action recognition datasets.The fusion process increases the length of feature vectors;therefore,improved whale optimization algorithm is proposed and selects the best features.The best selected features are finally classified usingmachine learning(ML)classifiers.Four publicly accessible datasets such as Ut-interaction,Hollywood,Free Viewpoint Action Recognition usingMotion History Volumes(IXMAS),and centre of computer vision(UCF)Sports,are employed and achieved the testing accuracy of 100%,99.9%,99.1%,and 100%respectively.Comparison with state of the art techniques(SOTA),the proposed method showed the improved accuracy.

Deep Attention Network for Pneumonia Detection Using Chest X-Ray Images

In computer vision,object recognition and image categorization have proven to be difficult challenges.They have,nevertheless,generated responses to a wide range of difficult issues from a variety of fields.Convolution Neural Networks(CNNs)have recently been identified as the most widely proposed deep learning(DL)algorithms in the literature.CNNs have unquestionably delivered cutting-edge achievements,particularly in the areas of image classification,speech recognition,and video processing.However,it has been noticed that the CNN-training assignment demands a large amount of data,which is in low supply,especially in the medical industry,and as a result,the training process takes longer.In this paper,we describe an attentionaware CNN architecture for classifying chest X-ray images to diagnose Pneumonia in order to address the aforementioned difficulties.AttentionModules provide attention-aware properties to the Attention Network.The attentionaware features of various modules alter as the layers become deeper.Using a bottom-up top-down feedforward structure,the feedforward and feedback attention processes are integrated into a single feedforward process inside each attention module.In the present work,a deep neural network(DNN)is combined with an attention mechanism to test the prediction of Pneumonia disease using chest X-ray pictures.To produce attention-aware features,the suggested networkwas built by merging channel and spatial attentionmodules in DNN architecture.With this network,we worked on a publicly available Kaggle chest X-ray dataset.Extensive testing was carried out to validate the suggested model.In the experimental results,we attained an accuracy of 95.47%and an F-score of 0.92,indicating that the suggested model outperformed against the baseline models.

Crops Leaf Diseases Recognition:A Framework of Optimum Deep Learning Features

Manual diagnosis of crops diseases is not an easy process;thus,a computerized method is widely used.Froma couple of years,advancements in the domain ofmachine learning,such as deep learning,have shown substantial success.However,they still faced some challenges such as similarity in disease symptoms and irrelevant features extraction.In this article,we proposed a new deep learning architecture with optimization algorithm for cucumber and potato leaf diseases recognition.The proposed architecture consists of five steps.In the first step,data augmentation is performed to increase the numbers of training samples.In the second step,pre-trained DarkNet19 deep model is opted and fine-tuned that later utilized for the training of fine-tuned model through transfer learning.Deep features are extracted from the global pooling layer in the next step that is refined using Improved Cuckoo search algorithm.The best selected features are finally classified using machine learning classifiers such as SVM,and named a few more for final classification results.The proposed architecture is tested using publicly available datasets–Cucumber National Dataset and Plant Village.The proposed architecture achieved an accuracy of 100.0%,92.9%,and 99.2%,respectively.Acomparison with recent techniques is also performed,revealing that the proposed method achieved improved accuracy while consuming less computational time.

Crystallography of low Z material at ultrahigh pressure:Case study on solid hydrogen

Diamond anvil cell techniques have been improved to allow access to the multimegabar ultrahigh-pressure region for exploring novel phenomena in condensedmatter.However,the onlyway to determine crystal structures of materials above 100 GPa,namely,X-ray diffraction(XRD),especially for lowZ materials,remains nontrivial in the ultrahigh-pressure region,even with the availability of brilliant synchrotron X-ray sources.In thiswork,we performa systematic study,choosing hydrogen(the lowest X-ray scatterer)as the subject,to understand how to better perform XRD measurements of low Z materials at multimegabar pressures.The techniques that we have developed have been proved to be effective in measuring the crystal structure of solid hydrogen up to 254GPa at room temperature[C.Ji et al.,Nature 573,558–562(2019)].Wepresent our discoveries and experienceswith regard to several aspects of thiswork,namely,diamond anvil selection,sample configuration for ultrahigh-pressure XRDstudies,XRDdiagnostics for low Z materials,and related issues in data interpretation and pressure calibration.Webelieve that these methods can be readily extended to other low Z materials and can pave the way for studying the crystal structure of hydrogen at higher pressures,eventually testing structural models of metallic hydrogen.

Evolution of Air Traffic Management Concept of Operations and Its Impact on the System Architecture

The air traffic management system(ATM)has the task of ensuring safe,orderly and expeditious flow of air traffic.The ATM system architecture is very much dependent on the concept of operations(ConOps).Over the years the evolution in ConOps has resulted in changes in the ATM′s physical architecture,improving its physical infrastructure,increasing the levels of automation and making operational changes to improve air traffic flow,to cope with increasing demand for air travel.However,what is less clear is the impact of such changes in ConOps on the ATM′s functional architecture.This is vital for ensuring optimality in the implementation of the physical architecture components to support the ATM functions.This paper reviews the changes in the ConOps over the years,proposes a temporally invariant ATM functional model,and discusses some of the main key technologies expected to make significant improvements to the ATM system.

Biomass estimation of Shorea robusta with principal component analysis of satellite data

Spatio-temporal assessment of the above ground biomass （AGB） is a cumbersome task due to the difficulties associated with the measurement of different tree parameters such as girth at breast height and height of trees. The present research was conducted in the campus of Birla Institute of Technology, Mesra, Ranchi, India, which is predomi- nantly covered by Sal （Shorea robusta C. F. Gaertn）. Two methods of regression analysis was employed to determine the potential of remote sensing parameters with the AGB measured in the field such as linear regression analysis between the AGB and the individual bands, principal components （PCs） of the bands, vegetation indices （VI）, and the PCs of the VIs respectively and multiple linear regression （MLR） analysis be- tween the AGB and all the variables in each category of data. From the linear regression analysis, it was found that only the NDVI exhibited regression coefficient value above 0.80 with the remaining parameters showing very low values. On the other hand, the MLR based analysis revealed significantly improved results as evidenced by the occurrence of very high correlation coefficient values of greater than 0.90 determined between the computed AGB from the MLR equations and field-estimated AGB thereby ascertaining their superiority in providing reliable estimates of AGB. The highest correlation coefficient of 0.99 is found with the MLR involving PCs of VIs.

Signet Ring Cell Detection from Histological Images Using Deep Learning

Signet Ring Cell(SRC)Carcinoma is among the dangerous types of cancers,and has a major contribution towards the death ratio caused by cancerous diseases.Detection and diagnosis of SRC carcinoma at earlier stages is a challenging,laborious,and costly task.Automatic detection of SRCs in a patient’s body through medical imaging by incorporating computing technologies is a hot topic of research.In the presented framework,we propose a novel approach that performs the identification and segmentation of SRCs in the histological images by using a deep learning(DL)technique named Mask Region-based Convolutional Neural Network(Mask-RCNN).In the first step,the input image is fed to Resnet-101 for feature extraction.The extracted feature maps are conveyed to Region Proposal Network(RPN)for the generation of the region of interest(RoI)proposals as well as they are directly conveyed to RoiAlign.Secondly,RoIAlign combines the feature maps with RoI proposals and generates segmentation masks by using a fully connected(FC)network and performs classification along with Bounding Box(bb)generation by using FC layers.The annotations are developed from ground truth(GT)images to perform experimentation on our developed dataset.Our introduced approach achieves accurate SRC detection with the precision and recall values of 0.901 and 0.897 respectively which can be utilized in clinical trials.We aim to release the employed database soon to assist the improvement in the SRC recognition research area.

Modeling the effects of stretch-dependent surfactant secretion on lung recruitment during variable ventilation

Variable ventilation (VV) is a novel strategy of ventilatory support that utilizes random variations in the delivered tidal volume (VT) to improve lung function. Since the stretch pattern during VV has been shown to increase surfactant release both in animals and cell culture, we hypothesized that there were combinations of PEEP and VT during VV that led to improved alveolar recruitment compared to conventional mechanical ventilation (CV). To test this hypothesis, we developed a computational model of stretch-induced surfactant release combined with abnormal alveolar mechanics of the injured lung under mechanical ventilation. We modeled the lung as a set of distinct acini with independent surfactant secretion and thus pressure-volume relationships. The rate of surfactant secretion was modulated by the stretch magnitude that an alveolus experienced per breath. Mechanical ventilation was simulated by delivering a prescribed VT at each breath. The fractional VT that each acinus received depended on its local compliance relative to the total system compliance. Regional variability in VT thus developed through feedback between stretch and surfactant release and coupling of regional VT to ventilator settings. The model allowed us to simulate patient-ventilator interactions over a wide range of PEEPs and VTs during CV and VV. Full recruitment was achieved through VV at a lower PEEP than required for CV. During VV, the acini were maintained under non-equilibrium steady-state conditions with breath-by-breath fluctuations of regional VT. In CV, alveolar injury was prevented with high-PEEP-low-VT or low-PEEP-high-VT combinations. In contrast, one contiguous region of PEEP-VT combinations allowed for full recruitment without overdistention during VV. We found that maintaining epithelial cell stretch above a critical threshold with either PEEP or VT may help stabilize the injured lung. These results demonstrate the significance of patient-ventilator coupling through the influence of cellular stretch-induced surfactant release on the whole lung stability.

Enhancing air traffic operational efficiency by reducing network scale

The growing demand for air travel has led to the saturation of air traffic networks.Conventional methods of adding routes to alleviate congestion and reduce delays may not achieve the desired effect and even degrade system performance.In this paper,we explore the application of Braess’s Paradox in the reduction of air traffic networks.This counterintuitive phenomenon shows that adding new connections to a network can actually increase the overall network pressure.This study uses Hidden Markov methods and the Viterbi algorithm to match air traffic flow with routes,a machine learning approach and a mathematical method to construct cost functions for flight time and traffic volume,and finally uses genetic algorithm and the A*algorithm to detect Braess’s Paradox edges.We uses ADS-B data from the busy month of July 2019 for a case study of the air traffic network over the UK airspace.The results show that Braess’s Paradox is also applicable to multi-flight level air route networks.Removing such network edges can improve system performance.In one day’s case,the total flight time of the day’s traffic volume decreased from 11509.24 minutes to 10459.97 minutes.This equates to an average savings of 4.99 minutes of flight time per flight,which is significant in controlling delay performance.

Comparative Assessment of the Relationship of Satellite Data with the Above Ground Biomass of Sal Trees(Shorea robusta) Determined from Phenologically Different Time Periods

The Above Ground Biomass(AGB) estimates of vegetation comprise both the bole biomass determined through a volumetric equation and litter biomass collected from the ground.For mature trees,the AGB estimated in phenologically different time periods is directly affected by the litter biomass since the Diameter at Breast Height(DBH) and height(H) of such trees that are used in the estimation of bole biomass would remain unchanged over a reasonable time period.In the present study,we have determined the AGB of Sal trees(Shorea robusta) in two contrasting seasons:the peak green period in October being devoid of lit-ter on the ground and the leaf shedding period in February with abundant amount of litter present on the ground.Estimation of AGB for the month of February included the litter biomass.In contrast,the AGB for October represented only the bole biomass.AGB was estimated for ten different plots selected in the study area.The AGB estimated from ten sampling plots for each time period was re-gressed with the individual tree parameters such as the average DBH and height of trees measured from the corresponding plots.The regression analysis exhibited a significantly stronger relationship between the AGB and DBH for the month of October as compared to February.Furthermore,the correlation between the remotely sensed derived data and AGB was also found to be significantly higher for the month of October than February.This observation indicates that inclusion of the litter biomass in AGB will tend to decrease the re-gression relationship between AGB and DBH and also between the remotely sensed data and AGB.Therefore,these conclusions invite careful consideration while estimating AGB from satellite data in phenologically different time periods.