Geo-Spatial Disease Clustering for Public Health Decision Making

Informatica, Journal Year: 2022, Volume and Issue: 46(6)

Published: Aug. 2, 2022

An explosion of interest has been observed in disease mapping with the developments advanced spatial statistics and increasing availability computerized geographic information system (GIS) technology. This technique is known as "Disease Clustering" using this for future prediction termed "Geo-Spatial Disease Clustering". Government, Medical Institutes, other medical practices gather large amounts data from surveys sources. form hard copies, databases, spread sheets text files. Mostly feedback different classes like age group, gender, provider (doctors), region, etc. During research used experiments testing. Variety techniques algorithms have proposed literature mapping. The effectiveness these may vary varying types, volume, structure interest. In research, investigation visualization proposed. includes cleansing, fusion, dimensioning, analysis, visualization, prediction. Motivation behind to create awareness about guidance patient healthcare providers government bodies. By this, we can extract that describes association respect age, location. Moreover, temporal analysis helps earlier identification disease, be care necessary avoiding arrangements taken.

Language: Английский

---

Multimodal Autism Spectrum Disorder Diagnosis Method Based on DeepGCN

IEEE Transactions on Neural Systems and Rehabilitation Engineering, Journal Year: 2023, Volume and Issue: 31, P. 3664 - 3674

Published: Jan. 1, 2023

Multimodal data play an important role in the diagnosis of brain diseases. This study constructs a whole-brain functional connectivity network based on MRI data, uses non-imaging with demographic information to complement classification task for diagnosing subjects, and proposes multimodal across-site WL-DeepGCN-based method diagnose autism spectrum disorder (ASD). is used resolve existing problem that deep learning ASD identification cannot efficiently utilize data. In WL-DeepGCN, weight-learning represent similarity latent space, introducing new approach constructing population graph edge weights, we find it beneficial robust define pairwise associations space rather than input space. We propose convolutional neural residual reduce loss due convolution operations by units avoid gradient disappearance explosion. Furthermore, EdgeDrop strategy makes node connections sparser randomly dropping edges raw graph, its introduction can alleviate overfitting oversmoothing problems DeepGCN training process. compare WL-DeepGCN model competitive models same topics nested 10-fold cross-validation show our achieves 77.27% accuracy 0.83 AUC identification, bringing substantial performance gains.

Language: Английский

---

Federated Machine Learning Based Fetal Health Prediction Empowered with Bio-Signal Cardiotocography

Computers, materials & continua/Computers, materials & continua (Print), Journal Year: 2024, Volume and Issue: 78(3), P. 3303 - 3321

Published: Jan. 1, 2024

Cardiotocography measures the fetal heart rate in fetus during pregnancy to ensure physical health because cardiotocography gives data about and uterine shrinkages which is very beneficial detect whether normal or suspect pathologic.Various infer wrongly give wrong predictions of human error.The traditional way reading time taken belongs numerous errors as well.Fetal condition important measure at stages proper medications for its well-being.In current period Machine learning (ML) a well-known classification strategy used biomedical field on various issues ML fast appropriate results that are better than results.ML techniques play pivotal role detecting disease early stages.This research article uses Federated machine (FML) classify fetus.This study proposed model detection bio-signal FML train test data.So, preprocessing overcome deficiency achieves 99.06% 0.94% prediction accuracy misprediction rate, respectively, parallel applying K-nearest neighbor (KNN) 82.93% 17.07% accuracy, respectively.So, by comparing both models outperformed KNN technique achieved best most compared with previous studies accurate results.

Language: Английский

---

A multi-view convolutional neural network method combining attention mechanism for diagnosing autism spectrum disorder

PLoS ONE, Journal Year: 2023, Volume and Issue: 18(12), P. e0295621 - e0295621

Published: Dec. 8, 2023

Autism Spectrum Disorder (ASD) is a neurodevelopmental condition whose current psychiatric diagnostic process subjective and behavior-based. In contrast, functional magnetic resonance imaging (fMRI) can objectively measure brain activity useful for identifying disorders. However, the ASD models employed to date have not reached satisfactory levels of accuracy. This study proposes use MAACNN, method that utilizes multi-view convolutional neural networks (CNNs) in conjunction with attention mechanisms multi-scale fMRI. The proposed algorithm effectively combines unsupervised supervised learning. initial stage, we employ stacked denoising autoencoders, an learning feature extraction, which provides different nodes adapt data. subsequent perform by employing CNNs classification obtain final results. Finally, data fusion achieved using mechanism. ABIDE dataset used evaluate model proposed., experimental results show MAACNN achieves superior performance 75.12% accuracy 0.79 AUC on ABIDE-I, 72.88% 0.76 ABIDE-II. significantly contributes clinical diagnosis ASD.

Language: Английский

---

Spam and Sentiment Detection in Arabic Tweets Using MARBERT Model

Mathematical Modelling and Engineering Problems, Journal Year: 2022, Volume and Issue: 9(6), P. 1574 - 1582

Published: Dec. 31, 2022

Saudi Telecom Company (STC) is among the most popular companies in Arabia, with many customers. Yet, there still a big room for improvement users' satisfaction. Social media robust platform to gauge satisfaction and determine their sentiments critics. Twitter social this regard. STC customers prefer use write feedback because it's fast way get responses due customer services account. One achieve demands improve service using Sentiment Analysis tool. on highly used of significant number tweets different opinions. Likewise, Deep learning best existing method, it has diverse models. Bidirectional Encoder Representations from Transformers (BERT) model one deep models which have achieved excellent results Natural Language Processing (NLP). NLP mainly investigated English language. However, Arabic, gap be filled. This study trained proposed MARBERT measured performance f1-score, precision, recall metrics. We an Arabic dataset 24,513 tweets, including 1,437 positive, 13,828 negative, 5,694 neutral, 1,221 sarcasm, 2,297 indeterminate tweets. The main goal analyze sentiment service. scheme promising terms accuracy contrast techniques literature.

Language: Английский

---

An Automated Platform for Gathering and Managing Open-Source Cyber Threat Intelligence

2022 International Conference on Business Analytics for Technology and Security (ICBATS), Journal Year: 2023, Volume and Issue: unknown, P. 1 - 7

Published: March 7, 2023

The community has begun paying more attention to source OSCTI Cyber Threat Intelligence stay informed about the rapidly changing cyber threat landscape. Numerous reports from frequently provide Information dangers. However, current gathering and management tools have mainly concentrated on individual minor compromise indicators, despite urgent need for high-quality OSCTI. relationship between higher-level notions (including strategies, methods, processes) connections them, which hold crucial dangerous behaviors are revealing full situation, been disregarded. Therefore, we present SecurityKG, an automated collection administration system. SecurityKG collects extract high-fidelity knowledge behaviours address void. Using a mixture of AI NLP approaches, security know-how graph is then constructed wide variety sources. To facilitate exploration, provides user interface (UI) that supports multiple forms interactivity.

Language: Английский

---

Heart Disease Prediction Using Machine Learning

2022 International Conference on Business Analytics for Technology and Security (ICBATS), Journal Year: 2023, Volume and Issue: unknown, P. 1 - 6

Published: March 7, 2023

The heart disease cases are rising day by and it is very Important to predict such diseases before causes more harm human lives. diagnosis of a complex task i.e., should be performed carefully. work done in this research paper mainly focuses on which patients has chance suffer from based their various medical feature as chest pain etc. We proposed system prediction that used diagnose whether the patient victim or not using previous features patient. Support vector machine k-nearest neighbor algorithms learning classify with disease. models gave satisfactory results were capable for predicting support good accuracy contrast naive bayes

Language: Английский

---

Dementia prediction with multimodal clinical and imaging data

International Journal of Information Technology, Journal Year: 2024, Volume and Issue: unknown

Published: Dec. 6, 2024

Language: Английский

---

A Review on the need of Clustering Techniques Used for Wireless Sensor Networks

2022 International Conference on Business Analytics for Technology and Security (ICBATS), Journal Year: 2023, Volume and Issue: unknown, P. 1 - 7

Published: March 7, 2023

Wireless sensor networks have been implemented in various software to help gather and analyze data from the physical world. are of multiple tiny sensors powered by low-energy batteries (WSNs). WSN lifetime is a critical factor. This because, after WSNs deployed given application, inaccessibility sensing nodes makes it impossible recharge or replace power source that limited energy. Until recently, were thought be same. All network same power, processing speed, operational capacity. However, scientists developed heterogeneous WSNs, which properties individual may vary, extend lifespan networks. In order network's helpful life, energy-efficient protocols must designed. Even though has its own problems, brought fresh perspective studying real-time intelligent systems. (Wireless Sensor Network) gained significant popularity recent years, owing vast applications scenarios, such as disaster management, pollution monitoring, temperature traffic transport healthcare battlefield border security surveillance, name few. Numerous employed these applications. They typically placed field, where they automatically transfer base station (BS) through energy transmission (i.e., battery sensor). WSN, clustering one method for optimizing consumption at node. The grouped leader when undergo clustering. study overviews algorithms, each categorized according properties.

Language: Английский

---

Optimised stacked machine learning algorithms for genomics and genetics disorder detection in the healthcare industry

Functional & Integrative Genomics, Journal Year: 2024, Volume and Issue: 24(1)

Published: Feb. 1, 2024

Language: Английский

---