Urban public health spatial planning using big data technology and visual communication in IoT

Mathematical Biosciences & Engineering, Journal Year: 2023, Volume and Issue: 20(5), P. 8583 - 8600

Published: Jan. 1, 2023

<abstract><p>The planning of urban public health spatial can not only help people's physical and mental but also to optimize protect the environment. It is great significance study methods spatial. The application effect traditional poor, in this paper, using big data technology visual communication Internet Things (IoT) proposed. First, architecture established IoT, which divided into perception layer, network layer layer; Second, information collection performed at used simplify model information, automatically sort out data, establish a space evaluation system according type characteristics data; Finally, planned based on assessment results design concept through layer. show that when number regions reaches 60,000, maximum time region merging 7.86s. percentage fitting error 0.17. height 0.31m. average deviation coordinates 0.23, realize different spaces.</p></abstract>

Language: Английский

---

Using Machine Learning Algorithms to Predict Patient Portal Use Among Emergency Department Patients With Diabetes Mellitus

Journal of Clinical Medicine Research, Journal Year: 2023, Volume and Issue: 15(3), P. 133 - 138

Published: March 1, 2023

Different machine learning (ML) technologies have been applied in healthcare systems with diverse applications. We aimed to determine the model feasibility and accuracy of predicting patient portal use among diabetic patients by using six different ML algorithms. In addition, we also compared performance only essential variables.This was a single-center retrospective observational study. From March 1, 2019 February 28, 2020, included all from study emergency department (ED). The primary outcome status use. A total 18 variables consisting sociodemographic characteristics, ED clinic information, medical conditions were predict Six algorithms (logistic regression, random forest (RF), deep forest, decision tree, multilayer perception, support vector machine) used for such predictions. During initial step, predictions performed variables. Then, chosen via feature selection. Patient repeated accuracies (overall accuracy, sensitivity, specificity, area under receiver operating characteristic curve (AUC)) compared.A 77,977 unique placed our final analysis. Among them, 23.4% (18,223) mellitus (DM). found 26.9% DM patients. Overall, above 80% five out RF outperformed others when (accuracy 0.9876, sensitivity 0.9454, specificity 0.9969, AUC 0.9712). When eight chosen, still 0.9374, 0.9932, 0.9769).It is possible outcomes are fair accuracy. However, similar prediction accuracies, selection techniques can improve interpretability addressing most relevant features.

Language: Английский

---

Detecting dengue fever in children using online Rasch analysis to develop algorithms for parents: An APP development and usability study

Medicine, Journal Year: 2023, Volume and Issue: 102(13), P. e33296 - e33296

Published: March 31, 2023

Dengue fever (DF) is a significant public health concern in Asia. However, detecting the disease using traditional dichotomous criteria (i.e., absent vs present) can be extremely difficult. Convolutional neural networks (CNNs) and artificial (ANNs), due to their use of large number parameters for modeling, have shown potential improve prediction accuracy (ACC). To date, there has been no research conducted understand item features responses online Rasch analysis. verify hypothesis that combination CNN, ANN, K-nearest-neighbor algorithm (KNN), logistic regression (LR) ACC DF children, further required.We extracted 19 feature variables related symptoms from 177 pediatric patients, whom 69 were diagnosed with DF. Using RaschOnline technique analysis, we examined 11 statistical significance predicting risk Based on 2 sets data, 1 training (80%) other testing (20%), calculated by comparing areas under receiver operating characteristic curve (AUCs) between + DF- both sets. In set, compared scenarios: combined scheme individual algorithms.Our findings indicate visual displays data are easily interpreted analysis; k-nearest neighbors lower AUC (<0.50); LR relatively higher (0.70); all 3 algorithms an almost equal (=0.68), which smaller than Naive Bayes, raw Bayes normalized data; developed app assist parents children during dengue season.The development LR-based APP detection completed. help family members, clinicians differentiate febrile illnesses at early stage, 11-item model proposed developing APP.

Language: Английский

---

IoT-Enabled Deep Learning Algorithm for Estimation of State-of-Charge of Lithium-ion Batteries

Journal of Circuits Systems and Computers, Journal Year: 2023, Volume and Issue: 33(07)

Published: Oct. 27, 2023

Battery Management System (BMS) functions to monitor individual cell in a battery pack and its crucial task is maintain stability throughout the pack. The BMS responsible for maintaining safety of as well not harm user or environment. parameters that are be monitored Voltage, Current Temperature. With collected data, carefully monitors charging–discharging behavior particularly Lithium-ion (Li-ion) batteries which charging discharging completely different. This paper proposes real-time IOT connected deep learning algorithm estimation State-of-Charge (SoC) Li-ion batteries. provides unique objectives congruence between model-based conventional methods state-of-the-art algorithm, specifically Feed Forward Neural Network (FNN) nonRecurrent. also highlights advantages Internet-of-Things (IoT) Hybrid Electric Vehicles (HEVs) (EVs). major advantage proposed method Artificial Intelligence (AI)-based techniques aim bring error less than 2% at low cost time without model battery, par with Extended Kalman Filter (EKF) best ever practical theory. Another an abnormal condition (i.e., Unsafe Temperature) IF Then That (IFTTT) IoT mobile application interfaced through ThingSpeak cloud, sends notification alert expert prior emergency. Finally, data cloud platform future research analysis.

Language: Английский

---

Cancer-associated fibroblasts (CAFs) based model reveals potential for predicting bladder cancer patients’ prognoses and immunotherapy responses

Journal of Cancer Metastasis and Treatment, Journal Year: 2024, Volume and Issue: unknown

Published: Feb. 1, 2024

Aim: The purpose of this study is to enhance the understanding bladder cancer and role cancer-associated fibroblasts (CAFs) in its progression. We aim identify CAF-specific biomarkers develop a prognostic prediction model based on CAFs, thereby contributing advancement treatment strategies identification predictive for cancer. Method: employed single-cell RNA sequencing detect CAFs cells. Bladder cohorts were categorized into low- high-CAF groups using ssGSEA algorithm. also explored association between CAF-related scores, immune-related cells, immune checkpoint-related genes. Furthermore, we performed GSVA analysis understand biological features their link various cancer-related pathways. Result: Ten genes identified as CAF markers A significant difference was found with 2712 differentially expressed low-CAF tissues. CAFs-based included nine (ALDH1L2, AL450384.2, EMP1, LINC02362, WFIKKN1, GOLGA8A, POU5F1, AL354919.2, PTPRR), which are potentially crucial predicting prognosis. revealed involvement several pathways, such WNT, toll-like receptor, TGF-beta, MAPK, MTOR signaling model. Conclusion: This highlights progression prognosis constructed provide valuable insights future research potential therapeutic targets. CAF-dependent pathways promising development new treatments improving patients.

Language: Английский

---