A novel meta learning based stacked approach for diagnosis of thyroid syndrome

PLoS ONE, Journal Year: 2024, Volume and Issue: 19(11), P. e0312313 - e0312313

Published: Nov. 1, 2024

Thyroid syndrome, a complex endocrine disorder, involves the dysregulation of thyroid gland, impacting vital physiological functions. Common causes include autoimmune disorders, iodine deficiency, and genetic predispositions. The effects syndrome extend beyond itself, affecting metabolism, energy levels, overall well-being. is associated with severe cases dysfunction, highlighting potentially life-threatening consequences untreated or inadequately managed disorders. This research aims to propose an advanced meta-learning approach for timely detection syndrome. We used standard thyroid-balanced dataset containing 7,000 patient records apply machine-learning methods. proposed novel model based on unique stack K-Neighbors (KN) Random Forest (RF) models. Then, Logistic Regression (LR) built collective experience stacked For first time, KRL (KN-RF-LR) method employed effective diagnosis Extensive experiments illustrated that outperformed state-of-the-art approaches, achieving impressive performance accuracy 98%. vindicated scores through k-fold cross-validation enhanced using hyperparameter tuning. Our revolutionized contributing enhancement human life by reducing mortality rates.

Language: Английский

---

IVNet: Transfer Learning Based Diagnosis of Breast Cancer Grading Using Histopathological Images of Infected Cells

IEEE Access, Journal Year: 2023, Volume and Issue: 11, P. 127880 - 127894

Published: Jan. 1, 2023

Breast cancer constitutes a significant global health concern that impacts millions of women across the world. The diagnosis breast involves categorizing grades based on histopathological characteristics tumor cells. While assessment remains established benchmark for diagnosis, it is hampered by time-consuming procedures, subjectivity, and susceptibility to human errors. This study introduces novel approach called ImageNet-VGG16 (IVNet) real-time within hospital environment. research experiments are conducted using dataset known as Jimma University Medical Center (JUMC) grading. Advanced image processing techniques applied preprocess data, enhancing performance. preprocessing utilization Holistically Nested Edge Detection (HED) Contrast Limited Adaptive Histogram Equalization (CLAHE) transformation stain normalization. We employ advanced neural network-based transfer learning analyze preprocessed images identify affected Various pre-trained models utilized, including convolutional networks (CNN) such VGG16, ResNet50, InceptionNetv3, ImageNet, MobileNetv3, EfficientNetV3, in comparative framework. principal objective this accurate classification into Grade-1, Grade-2 Grade-3. Through extensive experimental research, we achieved commendable 97% correct rate utilizing hybrid VGG16 ImageNet proposed feature engineering method, IVNet. also validate our performance other state-of-the-art data statistical t-test analysis. Furthermore, develop user-friendly Graphical User Interface (GUI) facilitates cell tracking images. Our diagnosing application offers valuable insights treatment planning assists medical professionals making prognoses. Moreover, can serve reliable decision support system pathologists clinicians, particularly settings constrained limited resources restricted access expertise equipment.

Language: Английский

---

MicrobeNet: An Automated Approach for Microbe Organisms Prediction Using Feature Fusion and Weighted CNN Model

International Journal of Computational Intelligence Systems, Journal Year: 2025, Volume and Issue: 18(1)

Published: March 13, 2025

Language: Английский

---

Optimized virtual reality design through user immersion level detection with novel feature fusion and explainable artificial intelligence

PeerJ Computer Science, Journal Year: 2024, Volume and Issue: 10, P. e2150 - e2150

Published: July 19, 2024

Virtual reality (VR) and immersive technology have emerged as powerful tools with numerous applications. VR creates a computer-generated simulation that immerses users in virtual environment, providing highly realistic interactive experience. This finds applications various fields, including gaming, healthcare, education, architecture, training simulations. Understanding user immersion levels is crucial challenging for optimizing the design of Immersion refers to extent which feel absorbed engrossed environment. research primarily aims detect using an efficient machine-learning model. We utilized benchmark dataset based on experiences environments conduct our experiments. Advanced deep machine learning approaches are applied comparison. proposed novel technique called Polynomial Random Forest (PRF) feature generation mechanisms. The PRF approach extracts polynomial class prediction probability features generate new set. Extensive experiments show random forest outperformed state-of-the-art approaches, achieving high level detection rate 98%, technique. hyperparameter optimization cross-validation validate performance scores. Additionally, we explainable artificial intelligence (XAI) interpret reasoning behind decisions made by model VR. Our has potential revolutionize VR, enhancing process.

Language: Английский

---

Deep Learning-Based Classification of Macrofungi: Comparative Analysis of Advanced Models for Accurate Fungi Identification

Sensors, Journal Year: 2024, Volume and Issue: 24(22), P. 7189 - 7189

Published: Nov. 9, 2024

This study focuses on the classification of six different macrofungi species using advanced deep learning techniques. Fungi species, such as Amanita pantherina, Boletus edulis, Cantharellus cibarius, Lactarius deliciosus, Pleurotus ostreatus and Tricholoma terreum were chosen based their ecological importance distinct morphological characteristics. The research employed 5 machine techniques 12 models, including DenseNet121, MobileNetV2, ConvNeXt, EfficientNet, swin transformers, to evaluate performance in identifying fungi from images. DenseNet121 model demonstrated highest accuracy (92%) AUC score (95%), making it most effective distinguishing between species. also revealed that transformer-based particularly transformer, less effective, suggesting room for improvement application this task. Further advancements could be achieved by expanding datasets, incorporating additional data types biochemical, electron microscopy, RNA/DNA sequences, ensemble methods enhance performance. findings contribute valuable insights into both use biodiversity conservation

Language: Английский

---

Are the soils degraded by the photovoltaic power plant?

Cogent Food & Agriculture, Journal Year: 2024, Volume and Issue: 10(1)

Published: Feb. 6, 2024

New photovoltaic panels are installed on agricultural land every day and yet their effect the quality of soil has not been fully verified. Unfortunately, there many scientific works that focus real in conditions. The presented work intended to establish basic principles through which placement changes surrounding soil. Since is a very complex system, six properties were worked on, labeled as 'master properties' by Kuzyakov Zamanian. It was found photovol power plants can have positive under certain According our conclusions, it be assumed PV will number properties, we mainly expect an increase stability aggregates, content organic matter increased development microbial community.

Language: Английский

---

Prediction of leukemia peptides using convolutional neural network and protein compositions

BMC Cancer, Journal Year: 2024, Volume and Issue: 24(1)

Published: July 26, 2024

Abstract Leukemia is a type of blood cell cancer that in the bone marrow’s blood-forming cells. Two types are acute and chronic; enhances fast chronic growth gradually which further classified into lymphocytic myeloid leukemias. This work evaluates unique deep convolutional neural network (CNN) classifier improves identification precision by carefully examining concatenated peptide patterns. The study uses leukemia protein expression for experiments supporting two different techniques including independence applied cross-validation. In addition to CNN, multilayer perceptron (MLP), gated recurrent unit (GRU), (RNN) applied. experimental results show CNN model surpasses competitors with its outstanding predictability independent cross-validation testing on features extracted from expressions such as amino acid composition (AAC) group AAC (GAAC), tripeptide (TPC) TPC (GTPC), dipeptide (DPC) calculating accuracies their receiver operating characteristic (ROC) curve. testing, feature GAAC using MLP modules, ROC curves achieved overall 100% accuracy detection patterns 98.33% highest module. Furthermore, 0.965% extraordinary result GRU findings excellent at figuring out illnesses higher accuracy.

Language: Английский

---

Microbial Taxonomy: An Artful Exploration of Microbes with Neural Networks

Published: Dec. 16, 2023

Language: Английский

---