Lung and Colon Cancer Classification Using Multiscale Deep Features Integration of Compact Convolutional Neural Networks and Feature Selection

Technologies, Journal Year: 2025, Volume and Issue: 13(2), P. 54 - 54

Published: Feb. 1, 2025

The automated and precise classification of lung colon cancer from histopathological photos continues to pose a significant challenge in medical diagnosis, as current computer-aided diagnosis (CAD) systems are frequently constrained by their dependence on singular deep learning architectures, elevated computational complexity, ineffectiveness utilising multiscale features. To this end, the present research introduces CAD system that integrates several lightweight convolutional neural networks (CNNs) with dual-layer feature extraction selection overcome aforementioned constraints. Initially, it extracts attributes two separate layers (pooling fully connected) three pre-trained CNNs (MobileNet, ResNet-18, EfficientNetB0). Second, uses benefits canonical correlation analysis for dimensionality reduction pooling layer reduce complexity. In addition, features encapsulate both high- low-level representations. Finally, benefit multiple network architectures while reducing proposed merges dual variables then applies variance (ANOVA) Chi-Squared most discriminative integrated CNN architectures. is assessed LC25000 dataset leveraging eight distinct classifiers, encompassing various Support Vector Machine (SVM) variants, Decision Trees, Linear Discriminant Analysis, k-nearest neighbours. experimental results exhibited outstanding performance, attaining 99.8% accuracy cubic SVM classifiers employing merely 50 ANOVA-selected features, exceeding performance individual markedly diminishing framework’s capacity sustain exceptional limited set renders especially advantageous clinical applications where diagnostic precision efficiency critical. These findings confirm efficacy multi-CNN, multi-layer methodology enhancing mitigating constraints systems.

Language: Английский

---

Enhancing affordable EEG to act as a quantitative EEG for inattention treatment using MATLAB

Neural Computing and Applications, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 6, 2025

Abstract Lack of attention is a chronic behavior in deficit hyperactivity disorder (ADHD), autism spectrum (ASD), and other disorders that harm academic social performance. ADHD whose typical symptoms include inattention, hyperactivity, impulsivity. They have major impact on the affected person’s function development. The electroencephalogram (EEG) device diagnostic tool, whereas quantitative EEG (QEEG) therapeutic tool for most mental disorders. QEEG applies neurofeedback method treatment. Neurofeedback technique training brain functions an alternative to traditional oral treatment inattention due its numerous side effects. proposed software can upgrade devices hospitals clinics into QEEGs capable neurofeedback. upgrading tools stages are introduced this study. cost 25 times less than purchase price device. (Open BCI) has been upgraded with MATLAB as system, integrating variety feature extraction methods detection such fractal dimension (FD), wavelet transform (WT), multi-resolution techniques (MR), empirical mode decomposition (EMD) which signified notable progress field. Furthermore, implemented easily customizable any forthcoming superior may arise. Earlier research distinguished differences between states relaxation concentration using simple fixed threshold. In paper, short utilized calculate adaptive thresholds optimize individual Different thresholding were employed EMD_Dt distinguish focused unfocused epochs. threshold results more accurate reaching benchmark 99.82%, opposed method, reaches accuracy 97.73%. findings assessed through pilot study involving 3483 epochs collected across 24 sessions from male female children aged 5 16. was evaluated be Specific, Measurable, Achievable, Realistic, Timed (SMART) effect size 0.85528336, significant.

Language: Английский

---

Comparison of Bioelectric Signals and Their Applications in Artificial Intelligence: A Review

Computers, Journal Year: 2025, Volume and Issue: 14(4), P. 145 - 145

Published: April 11, 2025

This review examines the role of various bioelectrical signals in conjunction with artificial intelligence (AI) and analyzes how these are utilized AI applications. The applications electroencephalography (EEG), electroretinography (ERG), electromyography (EMG), electrooculography (EOG), electrocardiography (ECG) diagnostic therapeutic systems focused on. Signal processing techniques discussed, relevant studies that have clinical research settings highlighted. Advances signal classification methodologies powered by significantly improved accuracy efficiency medical analysis. integration algorithms for real-time monitoring diagnosis, particularly personalized medicine, is emphasized. AI-driven approaches shown to potential enhance precision improve patient outcomes. However, further needed optimize models diverse environments fully exploit interaction between technologies.

Language: Английский

---

Cross-Context Stress Detection: Evaluating Machine Learning Models on Heterogeneous Stress Scenarios Using EEG Signals

AI, Journal Year: 2025, Volume and Issue: 6(4), P. 79 - 79

Published: April 14, 2025

Background/Objectives: This article addresses the challenge of stress detection across diverse contexts. Mental is a worldwide concern that substantially affects human health and productivity, rendering it critical research challenge. Although numerous studies have investigated through machine learning (ML) techniques, there has been limited on assessing ML models trained in one context utilized another. The objective ML-based systems to create generalize various Methods: study examines generalizability employing EEG recordings from two stress-inducing contexts: mental arithmetic evaluation (MAE) virtual reality (VR) gaming. We present data collection workflow publicly release portion dataset. Furthermore, we evaluate classical their generalizability, offering insights into influence training model performance, efficiency, related expenses. were acquired leveraging MUSE-STM hardware during stressful MAE VR gaming scenarios. methodology entailed preprocessing signals using wavelet denoising mother wavelets, individual aggregated sensor data, three models—linear discriminant analysis (LDA), support vector (SVM), K-nearest neighbors (KNN)—for classification purposes. Results: In Scenario 1, where was employed for testing, TP10 electrode attained an average accuracy 91.42% all classifiers participants, whereas SVM classifier achieved highest 95.76% participants. 2, adopting as testing maximum 88.05% with combination TP10, AF8, TP9 electrodes LDA peak 90.27% among optimal performance Symlets 4 Daubechies-2 Scenarios 1 respectively. Conclusions: results demonstrate although exhibit generalization capabilities stressors, significantly influenced by alignment between contexts, evidenced systematic cross-context evaluations 80/20 train–test split per participant quantitative metrics (accuracy, precision, recall, F1-score) averaged observed variations scenarios, classifiers, sensors provide empirical this claim.

Language: Английский

---

Improved ADHD Diagnosis Using EEG Connectivity and Deep Learning through Combining Pearson Correlation Coefficient and Phase-Locking Value

Neuroinformatics, Journal Year: 2024, Volume and Issue: 22(4), P. 521 - 537

Published: Oct. 18, 2024

Language: Английский

---

Comparative analysis of electroencephalogram (EEG) data gathered from the frontal region with other brain regions affected by attention deficit hyperactivity disorder (ADHD) through multiresolution analysis and machine learning techniques

Applied Neuropsychology Child, Journal Year: 2024, Volume and Issue: unknown, P. 1 - 15

Published: Oct. 1, 2024

Attention deficit hyperactivity disorder (ADHD) is a neurodevelopmental characterized by repeated patterns of hyperactivity, impulsivity, and inattention that limit daily functioning development. Electroencephalography (EEG) anomalies correspond to changes in brain connection activity. The authors propose utilizing empirical mode decomposition (EMD) discrete wavelet transform (DWT) for feature extraction machine learning (ML) algorithms categorize ADHD control subjects. For this study, the considered freely accessible data obtained from IEEE site. Studies have demonstrated range EEG patients, such as variations power spectra, coherence patterns, event-related potentials (ERPs). Some studies claimed brain's prefrontal cortex frontal regions collaborate intricate networks, disorders either them exacerbate symptoms ADHD. , Based on research ADHD, proposed study examines optimal position electrode identifying addition monitoring accuracy frontal/ other our also investigates groupings highest effect accurateness identification results demonstrate dataset classified with AdaBoost provided values accuracy, precision, specificity, sensitivity, F1-score 1.00, 0.70, 0.75, 0.71, respectively, whereas using random forest (RF) it 0.98, 0.64, 0.60, 0.81, detecting After detailed analysis, observed most accurate included all electrodes. believe processes can detect various problems children signals.

Language: Английский

---

DETEC-ADHD: A Data-Driven Web App for Early ADHD Detection Using Machine Learning and Electroencephalography

Big Data and Cognitive Computing, Journal Year: 2024, Volume and Issue: 9(1), P. 3 - 3

Published: Dec. 30, 2024

Attention Deficit Hyperactivity Disorder (ADHD) diagnosis is often challenging due to subjective assessments and symptom variability, which can delay accurate detection treatment. To address these limitations, this study introduces DETEC-ADHD, a web-based application that combines machine learning (ML) techniques with multi-source data enhance diagnostic accuracy. Unlike traditional approaches, DETEC-ADHD primarily utilizes extensive personal, medical, psychological information for its initial classification. further refines diagnoses by identifying ADHD subtypes (inattentive, hyperactive, combined) through theta/beta wave ratio analysis from EEG data, offering neurophysiological insights complement classification process. Logistic Regression, selected validated accuracy reliability, served as the ML model app. The case studies demonstrated DETEC-ADHD’s effectiveness, achieving 100% in children 90% adults. By integrating diverse sources real-time analysis, provides scalable, cost-effective, accessible solution subtype identification, addressing challenges supporting healthcare providers, particularly resource-limited environments.

Language: Английский

---

Unlocking the Potential of EEG: Convolutional Neural Networks for ADHD Classification

Published: July 27, 2024

Language: Английский

---