Data-Driven Cervical Cancer Prediction Model with Outlier Detection and Over-Sampling Methods

A Fuzzy Rule-Based System for Classification of Diabetes

Diabetes is a fatal disease that currently has no treatment. However, early diagnosis of diabetes aids patients to start timely treatment and thus reduces or eliminates the risk of severe complications. The prevalence of diabetes has been rising rapidly worldwide. Several methods have been introduced to diagnose diabetes at an early stage, however, most of these methods lack interpretability, due to which the diagnostic process cannot be explained. In this paper, fuzzy logic has been employed to develop an interpretable model and to perform an early diagnosis of diabetes. Fuzzy logic has been combined with the cosine amplitude method, and two fuzzy classifiers have been constructed. Afterward, fuzzy rules have been designed based on these classifiers. Lastly, a publicly available diabetes dataset has been used to evaluate the performance of the proposed fuzzy rule-based model. The results show that the proposed model outperforms existing techniques by achieving an accuracy of 96.47%. The proposed model has demonstrated great prediction accuracy, suggesting that it can be utilized in the healthcare sector for the accurate diagnose of diabetes.

Breakdown of a Nonlinear Stochastic Nipah Virus Epidemic Models through Efficient Numerical Methods

Background: Nipah virus (NiV) is a zoonotic virus (transmitted from animals to humans), which can also be transmitted through contaminated food or directly between people. According to a World Health Organization (WHO) report, the transmission of Nipah virus infection varies from animals to humans or humans to humans. The case fatality rate is estimated at 40% to 75%. The most infected regions include Cambodia, Ghana, Indonesia, Madagascar, the Philippines, and Thailand. The Nipah virus model is categorized into four parts: susceptible (S), exposed (E), infected (I), and recovered (R). Methods: The structural properties such as dynamical consistency, positivity, and boundedness are the considerable requirements of models in these fields. However, existing numerical methods like Euler-Maruyama and Stochastic Runge-Kutta fail to explain the main features of the biological problems. Results: The proposed stochastic non-standard finite difference (NSFD) employs standard and non-standard approaches in the numerical solution of the model, with positivity and boundedness as the characteristic determinants for efficiency and low-cost approximations. While the results from the existing standard stochastic methods converge conditionally or diverge in the long run, the solution by the stochastic NSFD method is stable and convergent over all time steps. Conclusions: The stochastic NSFD is an efficient, cost-effective method that accommodates all the desired feasible properties.

DeepDRG: Performance of Artificial Intelligence Model for Real-Time Prediction of Diagnosis-Related Groups

Nowadays, the use of diagnosis-related groups (DRGs) has been increased to claim reimbursement for inpatient care. The overall benefits of using DRGs depend upon the accuracy of clinical coding to obtain reasonable reimbursement. However, the selection of appropriate codes is always challenging and requires professional expertise. The rate of incorrect DRGs is always high due to the heavy workload, poor quality of documentation, and lack of computer assistance. We therefore developed deep learning (DL) models to predict the primary diagnosis for appropriate reimbursement and improving hospital performance. A dataset consisting of 81,486 patients with 128,105 episodes was used for model training and testing. Patients' age, sex, drugs, diseases, laboratory tests, procedures, and operation history were used as inputs to our multiclass prediction model. Gated recurrent unit (GRU) and artificial neural network (ANN) models were developed to predict 200 primary diagnoses. The performance of the DL models was measured by the area under the receiver operating curve, precision, recall, and F1 score. Of the two DL models, the GRU method, had the best performance in predicting the primary diagnosis (AUC: 0.99, precision: 83.2%, and recall: 66.0%). However, the performance of ANN model for DRGs prediction achieved AUC of 0.99 with a precision of 0.82 and recall of 0.57. The findings of our study show that DL algorithms, especially GRU, can be used to develop DRGs prediction models for identifying primary diagnosis accurately. DeepDRGs would help to claim appropriate financial incentives, enable proper utilization of medical resources, and improve hospital performance.

Machine learning-based statistical analysis for early stage detection of cervical cancer

Cervical cancer (CC) is the most common type of cancer in women and remains a significant cause of mortality, particularly in less developed countries, although it can be effectively treated if detected at an early stage. This study aimed to find efficient machine-learning-based classifying models to detect early stage CC using clinical data. We obtained a Kaggle data repository CC dataset which contained four classes of attributes including biopsy, cytology, Hinselmann, and Schiller. This dataset was split into four categories based on these class attributes. Three feature transformation methods, including log, sine function, and Z-score were applied to these datasets. Several supervised machine learning algorithms were assessed for their performance in classification. A Random Tree (RT) algorithm provided the best classification accuracy for the biopsy (98.33%) and cytology (98.65%) data, whereas Random Forest (RF) and Instance-Based K-nearest neighbor (IBk) provided the best performance for Hinselmann (99.16%), and Schiller (98.58%) respectively. Among the feature transformation methods, logarithmic gave the best performance for biopsy datasets whereas sine function was superior for cytology. Both logarithmic and sine functions performed the best for the Hinselmann dataset, while Z-score was best for the Schiller dataset. Various Feature Selection Techniques (FST) methods were applied to the transformed datasets to identify and prioritize important risk factors. The outcomes of this study indicate that appropriate system design and tuning, machine learning methods and classification are able to detect CC accurately and efficiently in its early stages using clinical data.

A Hybrid Method to Enhance Thick and Thin Vessels for Blood Vessel Segmentation

Retinal blood vessels have been presented to contribute confirmation with regard to tortuosity, branching angles, or change in diameter as a result of ophthalmic disease. Although many enhancement filters are extensively utilized, the Jerman filter responds quite effectively at vessels, edges, and bifurcations and improves the visualization of structures. In contrast, curvelet transform is specifically designed to associate scale with orientation and can be used to recover from noisy data by curvelet shrinkage. This paper describes a method to improve the performance of curvelet transform further. A distinctive fusion of curvelet transform and the Jerman filter is presented for retinal blood vessel segmentation. Mean-C thresholding is employed for the segmentation purpose. The suggested method achieves average accuracies of 0.9600 and 0.9559 for DRIVE and CHASE_DB1, respectively. Simulation results establish a better performance and faster implementation of the suggested scheme in comparison with similar approaches seen in the literature.

Automated Segmentation of Median Nerve in Dynamic Sonography Using Deep Learning: Evaluation of Model Performance

There is an emerging trend to employ dynamic sonography in the diagnosis of entrapment neuropathy, which exhibits aberrant spatiotemporal characteristics of the entrapped nerve when adjacent tissues move. However, the manual tracking of the entrapped nerve in consecutive images demands tons of human labors and impedes its popularity clinically. Here we evaluated the performance of automated median nerve segmentation in dynamic sonography using a variety of deep learning models pretrained with ImageNet, including DeepLabV3+, U-Net, FPN, and Mask-R-CNN. Dynamic ultrasound images of the median nerve at across wrist level were acquired from 52 subjects diagnosed as carpal tunnel syndrome when they moved their fingers. The videos of 16 subjects exhibiting diverse appearance and that of the remaining 36 subjects were used for model test and training, respectively. The centroid, circularity, perimeter, and cross section area of the median nerve in individual frame were automatically determined from the inferred nerve. The model performance was evaluated by the score of intersection over union (IoU) between the annotated and model-predicted data. We found that both DeepLabV3+ and Mask R-CNN predicted median nerve the best with averaged IOU scores close to 0.83, which indicates the feasibility of automated median nerve segmentation in dynamic sonography using deep learning.

Artificial intelligence in gynecologic cancers: Current status and future challenges - A systematic review

OBJECTIVE

Over the past years, the application of artificial intelligence (AI) in medicine has increased rapidly, especially in diagnostics, and in the near future, the role of AI in medicine will become progressively more important. In this study, we elucidated the state of AI research on gynecologic cancers.

METHODS

A search was conducted in three databases-PubMed, Web of Science, and Scopus-for research papers dated between January 2010 and December 2020. As keywords, we used "artificial intelligence," "deep learning," "machine learning," and "neural network," combined with "cervical cancer," "endometrial cancer," "uterine cancer," and "ovarian cancer." We excluded genomic and molecular research, as well as automated pap-smear diagnoses and digital colposcopy.

RESULTS

Of 1632 articles, 71 were eligible, including 34 on cervical cancer, 13 on endometrial cancer, three on uterine sarcoma, and 21 on ovarian cancer. A total of 35 studies (49%) used imaging data and 36 studies (51%) used value-based data as the input data. Magnetic resonance imaging (MRI), computed tomography (CT), ultrasound, cytology, and hysteroscopy data were used as imaging data, and the patients' backgrounds, blood examinations, tumor markers, and indices in pathological examination were used as value-based data. The targets of prediction were definitive diagnosis and prognostic outcome, including overall survival and lymph node metastasis. The size of the dataset was relatively small because 64 studies (90%) included less than 1000 cases, and the median size was 214 cases. The models were evaluated by accuracy scores, area under the receiver operating curve (AUC), and sensitivity/specificity. Owing to the heterogeneity, a quantitative synthesis was not appropriate in this review.

CONCLUSIONS

In gynecologic oncology, more studies have been conducted on cervical cancer than on ovarian and endometrial cancers. Prognoses were mainly used in the study of cervical cancer, whereas diagnoses were primarily used for studying ovarian cancer. The proficiency of the study design for endometrial cancer and uterine sarcoma was unclear because of the small number of studies conducted. The small size of the dataset and the lack of a dataset for external validation were indicated as the challenges of the studies.

Automated invasive cervical cancer disease detection at early stage through suitable machine learning model

Cervical cancer is a common cancer that affects women all over the world. This is the fourth leading cause of death among women and has no symptoms in its early stages. At the cervix, cervical cancer cells develop slowly. If it can be detected early, this cancer can be successfully treated. Health professionals are now facing a major challenge in detecting such cancer until it spreads rapidly. This study applied various machine learning classification methods to predict cervical cancer using risk factors. The main aim of this research work is to be described of the performance variation of eight most classifications algorithm to detect cervical cancer disease based on the selection of various top features sets from the dataset. Multilayer Perceptron (MLP), Random Forest and k-Nearest Neighbor, Decision Tree, Logistic Regression, SVC, Gradient Boosting, AdaBoost are examples of machine learning classification algorithms that have been used to predict cervical cancer and help in early diagnosis. A variety of approaches are used to avoid missing values in the dataset. To choose the various best features, a combination of feature selection techniques such as Chi-square, SelectBest and Random Forest was used. The performance of those classifications is evaluated using the accuracy, recall, precision and f1-score parameters. On a variety of top feature sets, MLP outperformed other classification models. The majority of classification models, on the other hand, claim to have the highest accuracy on the top 25 features in dataset splitting ratio (70:30). For each model, the percentage of correctly classified instances has been presented and all of the results are then discussed. Medical professionals will be able to use the suggested approach to perform research on cervical cancer.

Towards Interpretable Deep Learning: A Feature Selection Framework for Prognostics and Health Management Using Deep Neural Networks

In the last five years, the inclusion of Deep Learning algorithms in prognostics and health management (PHM) has led to a performance increase in diagnostics, prognostics, and anomaly detection. However, the lack of interpretability of these models results in resistance towards their deployment. Deep Learning-based models fall within the accuracy/interpretability tradeoff, which means that their complexity leads to high performance levels but lacks interpretability. This work aims at addressing this tradeoff by proposing a technique for feature selection embedded in deep neural networks that uses a feature selection (FS) layer trained with the rest of the network to evaluate the input features' importance. The importance values are used to determine which will be considered for deployment of a PHM model. For comparison with other techniques, this paper introduces a new metric called ranking quality score (RQS), that measures how performance evolves while following the corresponding ranking. The proposed framework is exemplified with three case studies involving health state diagnostics and prognostics and remaining useful life prediction. Results show that the proposed technique achieves higher RQS than the compared techniques, while maintaining the same performance level when compared to the same model but without an FS layer.

Multiclassification of Endoscopic Colonoscopy Images Based on Deep Transfer Learning

With the continuous improvement of human living standards, dietary habits are constantly changing, which brings various bowel problems. Among them, the morbidity and mortality rates of colorectal cancer have maintained a significant upward trend. In recent years, the application of deep learning in the medical field has become increasingly spread aboard and deep. In a colonoscopy, Artificial Intelligence based on deep learning is mainly used to assist in the detection of colorectal polyps and the classification of colorectal lesions. But when it comes to classification, it can lead to confusion between polyps and other diseases. In order to accurately diagnose various diseases in the intestines and improve the classification accuracy of polyps, this work proposes a multiclassification method for medical colonoscopy images based on deep learning, which mainly classifies the four conditions of polyps, inflammation, tumor, and normal. In view of the relatively small number of data sets, the network firstly trained by transfer learning on ImageNet was used as the pretraining model, and the prior knowledge learned from the source domain learning task was applied to the classification task about intestinal illnesses. Then, we fine-tune the model to make it more suitable for the task of intestinal classification by our data sets. Finally, the model is applied to the multiclassification of medical colonoscopy images. Experimental results show that the method in this work can significantly improve the recognition rate of polyps while ensuring the classification accuracy of other categories, so as to assist the doctor in the diagnosis of surgical resection.

Cytokine gene variants and socio-demographic characteristics as predictors of cervical cancer: A machine learning approach

Cervical cancer is still one of the most prevalent cancers in women and a significant cause of mortality. Cytokine gene variants and socio-demographic characteristics have been reported as biomarkers for determining the cervical cancer risk in the Indian population. This study was designed to apply a machine learning-based model using these risk factors for better prognosis and prediction of cervical cancer. This study includes the dataset of cytokine gene variants, clinical and socio-demographic characteristics of normal healthy control subjects, and cervical cancer cases. Different risk factors, including demographic details and cytokine gene variants, were analysed using different machine learning approaches. Various statistical parameters were used for evaluating the proposed method. After multi-step data processing and random splitting of the dataset, machine learning methods were applied and evaluated with 5-fold cross-validation and also tested on the unseen data records of a collected dataset for proper evaluation and analysis. The proposed approaches were verified after analysing various performance metrics. The logistic regression technique achieved the highest average accuracy of 82.25% and the highest average F1-score of 82.58% among all the methods. Ridge classifiers and the Gaussian Naïve Bayes classifier achieved the highest sensitivity-85%. The ridge classifier surpasses most of the machine learning classifiers with 84.78% accuracy and 97.83% sensitivity. The risk factors analysed in this study can be taken as biomarkers in developing a cervical cancer diagnosis system. The outcomes demonstrate that the machine learning assisted analysis of cytokine gene variants and socio-demographic characteristics can be utilised effectively for predicting the risk of developing cervical cancer.

Diagnosis of Cervical Cancer based on Ensemble Deep Learning Network using Colposcopy Images

Traditional screening of cervical cancer type classification majorly depends on the pathologist's experience, which also has less accuracy. Colposcopy is a critical component of cervical cancer prevention. In conjunction with precancer screening and treatment, colposcopy has played an essential role in lowering the incidence and mortality from cervical cancer over the last 50 years. However, due to the increase in workload, vision screening causes misdiagnosis and low diagnostic efficiency. Medical image processing using the convolutional neural network (CNN) model shows its superiority for the classification of cervical cancer type in the field of deep learning. This paper proposes two deep learning CNN architectures to detect cervical cancer using the colposcopy images; one is the VGG19 (TL) model, and the other is CYENET. In the CNN architecture, VGG19 is adopted as a transfer learning for the studies. A new model is developed and termed as the Colposcopy Ensemble Network (CYENET) to classify cervical cancers from colposcopy images automatically. The accuracy, specificity, and sensitivity are estimated for the developed model. The classification accuracy for VGG19 was 73.3%. Relatively satisfied results are obtained for VGG19 (TL). From the kappa score of the VGG19 model, we can interpret that it comes under the category of moderate classification. The experimental results show that the proposed CYENET exhibited high sensitivity, specificity, and kappa scores of 92.4%, 96.2%, and 88%, respectively. The classification accuracy of the CYENET model is improved as 92.3%, which is 19% higher than the VGG19 (TL) model.

A Dense Neural Network Approach for Detecting Clone ID Attacks on the RPL Protocol of the IoT

At present, new data sharing technologies, such as those used in the Internet of Things (IoT) paradigm, are being extensively adopted. For this reason, intelligent security controls have become imperative. According to good practices and security information standards, particularly those regarding security in depth, several defensive layers are required to protect information assets. Within the context of IoT cyber-attacks, it is fundamental to continuously adapt new detection mechanisms for growing IoT threats, specifically for those becoming more sophisticated within mesh networks, such as identity theft and cloning. Therefore, current applications, such as Intrusion Detection Systems (IDS), Intrusion Prevention Systems (IPS), and Security Information and Event Management Systems (SIEM), are becoming inadequate for accurately handling novel security incidents, due to their signature-based detection procedures using the matching and flagging of anomalous patterns. This project focuses on a seldom-investigated identity attack-the Clone ID attack-directed at the Routing Protocol for Low Power and Lossy Networks (RPL), the underlying technology for most IoT devices. Hence, a robust Artificial Intelligence-based protection framework is proposed, in order to tackle major identity impersonation attacks, which classical applications are prone to misidentifying. On this basis, unsupervised pre-training techniques are employed to select key characteristics from RPL network samples. Then, a Dense Neural Network (DNN) is trained to maximize deep feature engineering, with the aim of improving classification results to protect against malicious counterfeiting attempts.

Photoplethysmography in Normal and Pathological Sleep

This article presents an overview of the advancements that have been made in the use of photoplethysmography (PPG) for unobtrusive sleep studies. PPG is included in the quickly evolving and very popular landscape of wearables but has specific interesting properties, particularly the ability to capture the modulation of the autonomic nervous system during sleep. Recent advances have been made in PPG signal acquisition and processing, including coupling it with accelerometry in order to construct hypnograms in normal and pathologic sleep and also to detect sleep-disordered breathing (SDB). The limitations of PPG (e.g., oxymetry signal failure, motion artefacts, signal processing) are reviewed as well as technical solutions to overcome these issues. The potential medical applications of PPG are numerous, including home-based detection of SDB (for triage purposes), and long-term monitoring of insomnia, circadian rhythm sleep disorders (to assess treatment effects), and treated SDB (to ensure disease control). New contact sensor combinations to improve future wearables seem promising, particularly tools that allow for the assessment of brain activity. In this way, in-ear EEG combined with PPG and actigraphy could be an interesting focus for future research.

Voxel-Wise Feature Selection Method for CNN Binary Classification of Neuroimaging Data

Voxel-wise group analysis is presented as a novel feature selection (FS) technique for a deep learning (DL) approach to brain imaging data classification. The method, based on a voxel-wise two-sample t-test and denoted as t-masking, is integrated into the learning procedure as a data-driven FS strategy. t-Masking has been introduced in a convolutional neural network (CNN) for the test bench of binary classification of very-mild Alzheimer’s disease vs. normal control, using a structural magnetic resonance imaging dataset of 180 subjects. To better characterize the t-masking impact on CNN classification performance, six different experimental configurations were designed. Moreover, the performances of the presented FS method were compared to those of similar machine learning (ML) models that relied on different FS approaches. Overall, our results show an enhancement of about 6% in performance when t-masking was applied. Moreover, the reported performance enhancement was higher with respect to similar FS-based ML models. In addition, evaluation of the impact of t-masking on various selection rates has been provided, serving as a useful characterization for future insights. The proposed approach is also highly generalizable to other DL architectures, neuroimaging modalities, and brain pathologies.

Health-BlockEdge: Blockchain-Edge Framework for Reliable Low-Latency Digital Healthcare Applications

The rapid evolution of technology allows the healthcare sector to adopt intelligent, context-aware, secure, and ubiquitous healthcare services. Together with the global trend of an aging population, it has become highly important to propose value-creating, yet cost-efficient digital solutions for healthcare systems. These solutions should provide effective means of healthcare services in both the hospital and home care scenarios. In this paper, we focused on the latter case, where the goal was to provide easy-to-use, reliable, and secure remote monitoring and aid for elderly persons at their home. We proposed a framework to integrate the capabilities of edge computing and blockchain technology to address some of the key requirements of smart remote healthcare systems, such as long operating times, low cost, resilience to network problems, security, and trust in highly dynamic network conditions. In order to assess the feasibility of our approach, we evaluated the performance of our framework in terms of latency, power consumption, network utilization, and computational load, compared to a scenario where no blockchain was used.

VirtualCPR: Virtual Reality Mobile Application for Training in Cardiopulmonary Resuscitation Techniques

Deaths due to heart diseases are a leading cause of death in Mexico. Cardiovascular diseases are considered a public health problem because they produce cardiorespiratory arrests. During an arrest, cardiac and/or respiratory activity stops. A cardiorespiratory arrest is rapidly fatal without a quick and efficient intervention. As a response to this problem, the VirtualCPR system was designed in the present work. VirtualCPR is a mobile virtual reality application to support learning and practicing of basic techniques of cardiopulmonary resuscitation (CPR) for experts or non-experts in CPR. VirtualCPR implements an interactive virtual scenario with the user, which is visible by means of employment of virtual reality lenses. User’s interactions, with our proposal, are by a portable force sensor for integration with training mannequins, whose development is based on an application for the Android platform. Furthermore, this proposal integrates medical knowledge in first aid, related to the basic CPR for adults using only the hands, as well as technological knowledge, related to development of simulations on a mobile virtual reality platform by three main processes: (i) force measurement and conversion, (ii) data transmission and (iii) simulation of a virtual scenario. An experiment by means of a multifactorial analysis of variance was designed considering four factors for a CPR session: (i) previous training in CPR, (ii) frequency of compressions, (iii) presence of auditory suggestions and (iv) presence of color indicator. Our findings point out that the more previous training in CPR a user of the VirtualCPR system has, the greater the percentage of correct compressions obtained from a virtual CPR session. Setting the rate to 100 or 150 compressions per minute, turning on or off the auditory suggestions and turning the color indicator on or off during the session have no significant effect on the results obtained by the user.

An Examination System to Detect Deep Vein Thrombosis of a Lower Limb Using Light Reflection Rheography

Deep vein thrombosis (DVT) of lower limbs can easily arise from prolonged sitting or standing. Elders and pregnant women are most likely to have this disease. When the embolus of DVT comes to pass the lung, it will become a life-threatening disease. Thus, for DVT disease, early detection and the early treatment are needed. The goal of this study was to develop an examination system to be used at non-medical places to detect the DVT of lower limbs with light reflection rheography (LRR). Consisting of a wearable device and a mobile application (APP), the system is operated in a wireless manner to control the actions of sensors and display and store the LRR signals on the APP. Then, the recorded LRR signals are processed to find the parameters of DVT examination. Twenty subjects were recruited to perform experiments. The veins of lower limbs were occluded by pressuring the cuff up to 100 mmHg and 150 mmHg to simulate the slight and serious DVT scenarios, respectively. Six characteristic parameters were defined to classify whether there was positive or negative DVT using the receiver operating characteristic curves, including the slopes of emptying and refilling curves in the LRR signal, and the changes of venous pump volume. Under the slight DVT scenario (0 mmHg vs. 100 mmHg), the first three parameters, m₁₀, m₄₀, and m₅₀, had accuracies of 72%, 69%, and 69%, respectively. Under the serious DVT scenario (0 mmHg vs. 150 mmHg), m₁₀, m₄₀, and m₅₀ achieved accuracies of 73%, 76%, and 73%, respectively. The experimental results show that this proposed examination system may be practical as an auxiliary tool to screen DVT in homecare settings.

Generating Artificial Sensor Data for the Comparison of Unsupervised Machine Learning Methods

In the field of Cyber-Physical Systems (CPS), there is a large number of machine learning methods, and their intrinsic hyper-parameters are hugely varied. Since no agreed-on datasets for CPS exist, developers of new algorithms are forced to define their own benchmarks. This leads to a large number of algorithms each claiming benefits over other approaches but lacking a fair comparison. To tackle this problem, this paper defines a novel model for a generation process of data, similar to that found in CPS. The model is based on well-understood system theory and allows many datasets with different characteristics in terms of complexity to be generated. The data will pave the way for a comparison of selected machine learning methods in the exemplary field of unsupervised learning. Based on the synthetic CPS data, the data generation process is evaluated by analyzing the performance of the methods of the Self-Organizing Map, One-Class Support Vector Machine and Long Short-Term Memory Neural Net in anomaly detection.

Exposing Emerging Trends in Smart Sustainable City Research Using Deep Autoencoders-Based Fuzzy C-Means

The literature discussing the concepts, technologies, and ICT-based urban innovation approaches of smart cities has been growing, along with initiatives from cities all over the world that are competing to improve their services and become smart and sustainable. However, current studies that provide a comprehensive understanding and reveal smart and sustainable city research trends and characteristics are still lacking. Meanwhile, policymakers and practitioners alike need to pursue progressive development. In response to this shortcoming, this research offers content analysis studies based on topic modeling approaches to capture the evolution and characteristics of topics in the scientific literature on smart and sustainable city research. More importantly, a novel topic-detecting algorithm based on the deep learning and clustering techniques, namely deep autoencoders-based fuzzy C-means (DFCM), is introduced for analyzing the research topic trend. The topics generated by this proposed algorithm have relatively higher coherence values than those generated by previously used topic detection methods, namely non-negative matrix factorization (NMF), latent Dirichlet allocation (LDA), and eigenspace-based fuzzy C-means (EFCM). The 30 main topics that appeared in topic modeling with the DFCM algorithm were classified into six groups (technology, energy, environment, transportation, e-governance, and human capital and welfare) that characterize the six dimensions of smart, sustainable city research.

Modeling and Analyzing Offloading Strategies of IoT Applications over Edge Computing and Joint Clouds

Internet of Things (IoT) is swiftly evolving into a disruptive technology in recent years. For enhancing customer experience and accelerating job execution, IoT task offloading enables mobile end devices to release heavy computation and storage to the resource-rich nodes in collaborative Edges or Clouds. However, how different service architecture and offloading strategies quantitatively impact the end-to-end performance of IoT applications is still far from known particularly given a dynamic and unpredictable assortment of interconnected virtual and physical devices. This paper exploits potential network performance that manifests within the edge-cloud environment, then investigates and compares the impacts of two types of architectures: Loosely-Coupled (LC) and Orchestrator-Enabled (OE). Further, it introduces three customized offloading strategies in order to handle various requirements for IoT latency-sensitive applications. Through comparative experiments, we observed that the computational requirements exerts more influence on the IoT application’s performance compared to the communication requirement. However, when the system scales up to accommodate more IoT devices, communication bandwidth will turn to be the dominant resource and becomes the essential factor that will directly impact the overall performance. Thus, orchestration is a necessary procedure to encompass optimized solutions under different constraints for optimal offloading placement.

Predicting the Appearance of Hypotension During Hemodialysis Sessions Using Machine Learning Classifiers

A patient suffering from advanced chronic renal disease undergoes several dialysis sessions on different dates. Several clinical parameters are monitored during the different hours of any of these sessions. These parameters, together with the information provided by other parameters of analytical nature, can be very useful to determine the probability that a patient may suffer from hypotension during the session, which should be specially watched since it represents a proven factor of possible mortality. However, the analytical information is not always available to the healthcare personnel, or it is far in time, so the clinical parameters monitored during the session become key to the prevention of hypotension. This article presents an investigation to predict the appearance of hypotension during a dialysis session, using predictive models trained from a large dialysis database, which contains the clinical information of 98,015 sessions corresponding to 758 patients. The prediction model takes into account up to 22 clinical parameters measured five times during the session, as well as the gender and age of the patient. This model was trained by means of machine learning classifiers, providing a success in the prediction higher than 80%.

Brain Asymmetry Detection and Machine Learning Classification for Diagnosis of Early Dementia

Early identification of degenerative processes in the human brain is considered essential for providing proper care and treatment. This may involve detecting structural and functional cerebral changes such as changes in the degree of asymmetry between the left and right hemispheres. Changes can be detected by computational algorithms and used for the early diagnosis of dementia and its stages (amnestic early mild cognitive impairment (EMCI), Alzheimer’s Disease (AD)), and can help to monitor the progress of the disease. In this vein, the paper proposes a data processing pipeline that can be implemented on commodity hardware. It uses features of brain asymmetries, extracted from MRI of the Alzheimer’s Disease Neuroimaging Initiative (ADNI) database, for the analysis of structural changes, and machine learning classification of the pathology. The experiments provide promising results, distinguishing between subjects with normal cognition (NC) and patients with early or progressive dementia. Supervised machine learning algorithms and convolutional neural networks tested are reaching an accuracy of 92.5% and 75.0% for NC vs. EMCI, and 93.0% and 90.5% for NC vs. AD, respectively. The proposed pipeline offers a promising low-cost alternative for the classification of dementia and can be potentially useful to other brain degenerative disorders that are accompanied by changes in the brain asymmetries.

NADAL: A Neighbor-Aware Deep Learning Approach for Inferring Interpersonal Trust Using Smartphone Data

Interpersonal trust mediates multiple socio-technical systems and has implications for personal and societal well-being. Consequently, it is crucial to devise novel machine learning methods to infer interpersonal trust automatically using mobile sensor-based behavioral data. Considering that social relationships are often affected by neighboring relationships within the same network, this work proposes using a novel neighbor-aware deep learning architecture (NADAL) to enhance the inference of interpersonal trust scores. Based on analysis of call, SMS, and Bluetooth interaction data from a one-year field study involving 130 participants, we report that: (1) adding information about neighboring relationships improves trust score prediction in both shallow and deep learning approaches; and (2) a custom-designed neighbor-aware deep learning architecture outperforms a baseline feature concatenation based deep learning approach. The results obtained at interpersonal trust prediction are promising and have multiple implications for trust-aware applications in the emerging social internet of things.

Integrating Enhanced Sparse Autoencoder-Based Artificial Neural Network Technique and Softmax Regression for Medical Diagnosis

In recent times, several machine learning models have been built to aid in the prediction of diverse diseases and to minimize diagnostic errors made by clinicians. However, since most medical datasets seem to be imbalanced, conventional machine learning algorithms tend to underperform when trained with such data, especially in the prediction of the minority class. To address this challenge and proffer a robust model for the prediction of diseases, this paper introduces an approach that comprises of feature learning and classification stages that integrate an enhanced sparse autoencoder (SAE) and Softmax regression, respectively. In the SAE network, sparsity is achieved by penalizing the weights of the network, unlike conventional SAEs that penalize the activations within the hidden layers. For the classification task, the Softmax classifier is further optimized to achieve excellent performance. Hence, the proposed approach has the advantage of effective feature learning and robust classification performance. When employed for the prediction of three diseases, the proposed method obtained test accuracies of 98%, 97%, and 91% for chronic kidney disease, cervical cancer, and heart disease, respectively, which shows superior performance compared to other machine learning algorithms. The proposed approach also achieves comparable performance with other methods available in the recent literature.

P-NUT: Predicting NUTrient Content from Short Text Descriptions

Assessing nutritional content is very relevant for patients suffering from various diseases, professional athletes, and for health reasons is becoming part of everyday life for many. However, it is a very challenging task as it requires complete and reliable sources. We introduce a machine learning pipeline for predicting macronutrient values of foods using learned vector representations from short text descriptions of food products. On a dataset used from health specialists, containing short descriptions of foods and macronutrient values: we generate paragraph embeddings, introduce clustering in food groups, using graph-based vector representations, that include food domain knowledge information, and train regression models for each cluster. The predictions are for four macronutrients: carbohydrates, fat, protein and water. The highest accuracy was obtained for carbohydrate predictions – 86%, compared to the baseline – 27% and 36%. The protein predictions yielded the best results across all clusters, 53%–77% of the values fall in the tolerance-level range. These results were obtained using short descriptions, the embeddings can be improved if they are learned on longer descriptions, which would lead to better prediction results. Since the task of calculating macronutrients requires exact quantities of ingredients, these results obtained only from short description are a huge leap forward.

TADILOF: Time Aware Density-Based Incremental Local Outlier Detection in Data Streams

Outlier detection in data streams is crucial to successful data mining. However, this task is made increasingly difficult by the enormous growth in the quantity of data generated by the expansion of Internet of Things (IoT). Recent advances in outlier detection based on the density-based local outlier factor (LOF) algorithms do not consider variations in data that change over time. For example, there may appear a new cluster of data points over time in the data stream. Therefore, we present a novel algorithm for streaming data, referred to as time-aware density-based incremental local outlier detection (TADILOF) to overcome this issue. In addition, we have developed a means for estimating the LOF score, termed "approximate LOF," based on historical information following the removal of outdated data. The results of experiments demonstrate that TADILOF outperforms current state-of-the-art methods in terms of AUC while achieving similar performance in terms of execution time. Moreover, we present an application of the proposed scheme to the development of an air-quality monitoring system.

Validity Evaluation Method Based on Data Driving for On-Line Monitoring Data of Transformer under DC-Bias

The DC-bias monitoring device of a transformer is easily affected by external noise interference, equipment aging, and communication failure, which makes it difficult to guarantee the validity of monitoring data and causes great problems for future data analysis. For this reason, this paper proposes a validity evaluation method based on data driving for the on-line monitoring data of a transformer under DC-bias. First, the variation rule and threshold range of monitoring data for neutral point DC, vibration, and noise of the transformer under different working conditions are obtained through statistical analysis. Then, the data validity criterion of DC bias monitoring data is proposed to achieve a comprehensive evaluation of data validity based on data threshold, continuity, impact, and correlation. In addition, case studies are carried out on the real measured data of the DC bias magnetic monitoring system of a regional power grid by using this evaluation method. The results show that the proposed method can systematically and comprehensively evaluate the validity of the DC bias monitoring data and can judge whether the monitoring device fails to a certain extent.

A New Under-Sampling Method to Face Class Overlap and Imbalance

Class overlap and class imbalance are two data complexities that challenge the design of effective classifiers in Pattern Recognition and Data Mining as they may cause a significant loss in performance. Several solutions have been proposed to face both data difficulties, but most of these approaches tackle each problem separately. In this paper, we propose a two-stage under-sampling technique that combines the DBSCAN clustering algorithm to remove noisy samples and clean the decision boundary with a minimum spanning tree algorithm to face the class imbalance, thus handling class overlap and imbalance simultaneously with the aim of improving the performance of classifiers. An extensive experimental study shows a significantly better behavior of the new algorithm as compared to 12 state-of-the-art under-sampling methods using three standard classification models (nearest neighbor rule, J48 decision tree, and support vector machine with a linear kernel) on both real-life and synthetic databases.