Deep learning approaches to detect breast cancer: a comprehensive review

Multimedia Tools and Applications, Год журнала: 2024, Номер unknown

Опубликована: Авг. 20, 2024

Язык: Английский

---

Classification of α-thalassemia data using machine learning models

Computer Methods and Programs in Biomedicine, Год журнала: 2025, Номер 260, С. 108581 - 108581

Опубликована: Янв. 6, 2025

Around 7% of the global population has congenital hemoglobin disorders, with over 300,000 new cases α-thalassemia annually. Diagnosis is costly and inaccurate in low-income regions, often relying on complete blood count (CBC) tests. This study employs machine learning (ML) to classify traits based gender CBC, exploring effects grouping silent- non-carriers. The dataset includes 288 individuals suspected from Sri Lanka. It was classified using eleven discriminant formulae nine ML models. Outliers were removed Mahalanobis distance, resampling conducted synthetic minority oversampling technique (SMOTE) SMOTE-nominal continuous (NC). Mann-Whitney U test handled feature extraction class grouping. performance evaluated eight criteria. Ehsani formula achieved an area under receiver operating characteristic curve (ROC-AUC) 0.66 by convolutional neural network (CNN) without demonstrated better performance, accuracy 0.85, sensitivity 0.8, specificity 0.86, ROC-AUC 0.95/0.93 (micro/macro). Performance maintained even preprocessing. models outperformed classical classifying sex CBC features. A larger could enhance model generalization impact extraction. Grouping non-carriers improved results, especially resampling. silent carriers not separable regarding available

Язык: Английский

---

Deep ensemble learning for gastrointestinal diagnosis using endoscopic image classification

PeerJ Computer Science, Год журнала: 2025, Номер 11, С. e2809 - e2809

Опубликована: Апрель 22, 2025

Transfer learning is a valuable tool for the effective assistance of gastroenterologists in powerful diagnosis medical images with fast convergence. It also intends to minimize time and estimated effort required improved gastrointestinal tract (GIT) diagnosis. GIT abnormalities are widely known be fatal disorders leading significant mortalities. includes both upper lower disorders. The challenges addressing issues complex need study. Multiple exist regarding computer-aided (CAD) endoscopy including lack annotated images, dark backgrounds, less contrast, noisy irregular patterns. Deep transfer have assisted various ways. goal proposed framework classification endoscopic enhanced accuracy. research aims formulate learning-based deep ensemble model, accurately classifying therapeutic purposes. model based on weighted voting two state-of-the-art (STA) base models, NasNet-Mobile EfficientNet. extraction regions interest, specifically sick portions, been performed using captured from procedure. Performance evaluation cross-dataset evaluation. datasets utilized include training dataset HyperKvasir test datasets, Kvasir v1 v2. However, alone cannot create robust due unequal distribution across categories, making promising approach development. has conducted by utilizing accuracy, precision, recall, Area under curve (AUC) score F1 performance metrics. work outperforms much existing models giving 97.83% 98.45% accuracy

Язык: Английский

---

Evaluating EfficientNet Architectures for Pathology Detection in Endoscopic Gastrointestinal Tract Images

SN Computer Science, Год журнала: 2025, Номер 6(5)

Опубликована: Май 2, 2025

Язык: Английский

---

An Efficient Optimal CapsNet Model-Based Computer-Aided Diagnosis for Gastrointestinal Cancer Classification

IEEE Access, Год журнала: 2024, Номер 12, С. 137237 - 137246

Опубликована: Янв. 1, 2024

Язык: Английский

---

Classification of Gastrointestinal Diseases in Endoscopic Images: Comparative Analysis of Convolutional Neural Networks and Vision Transformers

Iğdır Üniversitesi Fen Bilimleri Enstitüsü Dergisi, Год журнала: 2024, Номер 14(3), С. 988 - 999

Опубликована: Авг. 17, 2024

Gastrointestinal (GI) diseases are a major issue in the human digestive system. Therefore, many studies have explored automatic classification of GI to reduce burden on clinicians and improve patient outcomes for both diagnosis treatment purposes. Convolutional neural networks (CNNs) Vision Transformers (ViTs) deep learning approaches become popular research area detection from medical images. This study evaluated performance thirteen different CNN models two ViT architectures endoscopic The impact transfer parameters was also observed. tests revealed that accuracies were 91.25% 90.50%, respectively. In contrast, DenseNet201 architecture, with optimized parameters, achieved an accuracy 93.13%, recall 93.17%, precision F1 score 93.11%, making it most successful model among all others. Considering results, is evident well-optimized better than models.

Язык: Английский

---

An Interpretable Glaucoma Detection using Dual Scale Cross Attention Vision Transformer-based Long Short Term Memory with Optical Cup and Disk Segmentation

Journal of Mechanics in Medicine and Biology, Год журнала: 2024, Номер unknown

Опубликована: Июль 18, 2024

Glaucoma is a kind of eye disease that tends to generate harm the optic nerve. It neurodegenerative illness, which develops intraocular hypertension because its maximized aqueous humor and blockage between cornea iris. causes destruction nerve head, transfers visual stimulus brain from eyes. This results in loss field blindness. For vision, glaucoma known be sneak thief due complexity detecting it early stage. requires continuous screening determine neurological disorder. Effective identification more cost time, but also human error detection phase based on resource availability. The problems robustness algorithm are not solved earlier method especially relative expert counterpart. Therefore, effective with help deep learning developed recognize At first, input images taken available sites. Subsequently, procedure for segmentation done using Optimized Dilated Mobile-Unet[Formula: see text] (ODMUnet[Formula: text]) segment disc cup images. Here, parameters ODMUnet[Formula: optimized an Improved Drawer Algorithm (IDrA). segmented “optic cup” given Dual Scale Cross-Attention Vision Transformer-based Long Short-Term Memory (DSCAViT-LSTM) detection. experimental outcomes recommended model evaluated other techniques ensure efficacy.

Язык: Английский

---

Lightweight Deep Learning Model Optimization for Medical Image Analysis

International Journal of Imaging Systems and Technology, Год журнала: 2024, Номер 34(5)

Опубликована: Сен. 1, 2024

ABSTRACT Medical image labeling requires specialized knowledge; hence, the solution to challenge of medical classification lies in efficiently utilizing few labeled samples create a high‐performance model. Building model complicated convolutional neural network (CNN) with numerous parameters be trained which makes test quite expensive. In this paper, we propose optimizing lightweight deep learning only five layers using particle swarm optimization (PSO) algorithm find best number kernel filters for each layer. For colored red, green, and blue (RGB) images acquired from different data sources, suggest stain separation color deconvolution horizontal vertical flipping produce new versions that can concentrate representation on structures patterns. To mitigate effect training incorrectly or uncertainly images, grades disease could have small variances, apply second‐pass excluding uncertain data. With higher accuracy, proposed (LDLMO) shows strong resilience generalization ability compared most recent research four MedMNIST datasets (RetinaMNIST, BreastMNIST, DermMNIST, OCTMNIST), Medical‐MNIST, brain tumor MRI datasets.

Язык: Английский

---

An efficient enhanced stacked auto encoder assisted optimized deep neural network for forecasting Dry Eye Disease

Scientific Reports, Год журнала: 2024, Номер 14(1)

Опубликована: Окт. 22, 2024

Meibomian Gland Dysfunction (MGD) and Dry Eye Disease (DED) comprise two of the most significant eye diseases, impacting millions sufferers worldwide. Several etiological factors influence early symptoms DED. Early diagnosis treatment erectile dysfunction may significantly improve Quality Life (QoL) for people. The current study introduces ESAE-ODNN, an improved stacked autoencoder-aided optimised deep neural network, as a new way to predict DED using feature selection (FS), extraction (FE), classification. approach described here is novel because it merges chaotic maps into FS, employs SLSTM-STSA classification accuracy (CA), optimizes with adaptive quantum rotation Enhanced Quantum Bacterial Foraging Optimisation Algorithm (EQBFOA). present enhances prediction functions by extracting MGD-related features complicated relationships from dataset. To ensure essential identification, ESAE minimizes irrelevant redundant features. DED, first applies FE then implements ODNN classifier. This method fine-tunes framework enhance effectiveness proposed ESAE-ODNN system efficiently assists in Combining advanced Deep Learning (DL) methods optimization can help us understand MGD better sort data best (96.34%). experimental evaluation relevant performance metrics indicates that efficient diverse aspects: accurate reduced complexity, fine-tuned performance. ESAE-ODNN's robustness handling intricate indications high-dimensional outperforms existing state-of-the-art techniques.

Язык: Английский

---

Enhancing Epstein–Barr virus detection in IBD patients with XAI and clinical data integration

Computers in Biology and Medicine, Год журнала: 2024, Номер 184, С. 109465 - 109465

Опубликована: Ноя. 22, 2024

Язык: Английский

---