Traditional machine learning algorithms for breast cancer image classification with optimized deep features

Biomedical Signal Processing and Control, Journal Year: 2022, Volume and Issue: 81, P. 104534 - 104534

Published: Dec. 22, 2022

Language: Английский

---

Boosted Nutcracker optimizer and Chaos Game Optimization with Cross Vision Transformer for medical image classification

Egyptian Informatics Journal, Journal Year: 2024, Volume and Issue: 26, P. 100457 - 100457

Published: April 9, 2024

This paper presents an alternative breast cancer classification method based on enhancing the efficiency of Nutcracker optimizer (NO) algorithm using Chaos Game Optimization (CGO). In addition, we use Cross Vision Transformer to extract features from images. After that, relevant are allocated modified version NO CGO. modification aims enhance exploration ability discover region a feasible solution (an optimal subset features). The performance developed model is validated by twelve functions CEC2022 benchmark and comparing results with traditional CGO algorithms. assess applicability technique, set three datasets, were compared other techniques. illustrate high detection find according different measures.

Language: Английский

---

Integrated EfficientNetB3V2 fused MaxEnt classifier model for brain tumor classification in MR images

Journal of the Chinese Institute of Engineers, Journal Year: 2025, Volume and Issue: unknown, P. 1 - 14

Published: Feb. 16, 2025

Language: Английский

---

Dual-Stream Contrastive Latent Learning Generative Adversarial Network for Brain Image Synthesis and Tumor Classification

Journal of Imaging, Journal Year: 2025, Volume and Issue: 11(4), P. 101 - 101

Published: March 28, 2025

Generative adversarial networks (GANs) prioritize pixel-level attributes over capturing the entire image distribution, which is critical in synthesis. To address this challenge, we propose a dual-stream contrastive latent projection generative network (DSCLPGAN) for robust augmentation of MRI images. The generator our architecture incorporates two specialized processing pathways: one dedicated to local feature variation modeling, while other captures global structural transformations, ensuring more comprehensive synthesis medical We used transformer-based encoder-decoder framework contextual coherence and learning (CLP) module integrates loss into space generating diverse samples. generated images undergo refinement using an ensemble discriminators, where discriminator 1 (D1) ensures classification consistency with real images, 2 (D2) produces probability map localized variations, 3 (D3) preserves consistency. For validation, utilized publicly available dataset contains 3064 T1-weighted contrast-enhanced three types brain tumors: meningioma (708 slices), glioma (1426 pituitary tumor (930 slices). experimental results demonstrate state-of-the-art performance, achieving SSIM 0.99, accuracy 99.4% diversity level 5, PSNR 34.6 dB. Our approach has potential high-fidelity augmentations reliable AI-driven clinical decision support systems.

Language: Английский

---

Brain tumor detection through novel feature selection using deep efficientNet-CNN-based features with supervised learning and data augmentation

Physica Scripta, Journal Year: 2024, Volume and Issue: 99(7), P. 075002 - 075002

Published: May 22, 2024

Abstract Brain tumors being ninth in terms of prevalence and one the most frequently diagnosed malignant tumors, negatively impact millions individuals. Identifying classifying from MRI used for health monitoring poses a challenge radiologists, yet early detection could significantly enhance chances effective treatment. Researchers field explainable AI are currently focused on developing sophisticated techniques to classify diagnose brain diseases. This study presents novel framework that enhances interpretability our proposed system tumor by utilizing techniques. To interpretability, we integrate optimized recursive feature elimination selection technique with support vector machines. method effectively eliminates redundant features, identifies important ones, efficiency detecting tasks. Following that, optimal (ORFE) features combined using supervised machine (SVM) technique. While EfficientNet-CNN is very useful extraction extracts transparent model, reduced overall computational complexity through models, Figshre dataset clearly demonstrated efficacy model. achieved exceptional results as compared single CNN The experimental indicate SVM-RFE based accurately detects 99.51% accuracy specificity score 99.63%. approach obtained an 98.93% standard deviation 0.032 10-fold cross-validation. Additionally, it produced ROC_AUC 100% cases including meningiomas pituitary tumors.

Language: Английский

---

Enhanced brain tumor detection and segmentation using densely connected convolutional networks with stacking ensemble learning

Computers in Biology and Medicine, Journal Year: 2025, Volume and Issue: 186, P. 109703 - 109703

Published: Jan. 24, 2025

Language: Английский

---

Precision meets generalization: Enhancing brain tumor classification via pretrained DenseNet with global average pooling and hyperparameter tuning

PLoS ONE, Journal Year: 2024, Volume and Issue: 19(9), P. e0307825 - e0307825

Published: Sept. 6, 2024

Brain tumors pose significant global health concerns due to their high mortality rates and limited treatment options. These tumors, arising from abnormal cell growth within the brain, exhibits various sizes shapes, making manual detection magnetic resonance imaging (MRI) scans a subjective challenging task for healthcare professionals, hence necessitating automated solutions. This study investigates potential of deep learning, specifically DenseNet architecture, automate brain tumor classification, aiming enhance accuracy generalizability clinical applications. We utilized Figshare dataset, comprising 3,064 T1-weighted contrast-enhanced MRI images 233 patients with three prevalent types: meningioma, glioma, pituitary tumor. Four pre-trained learning models—ResNet, EfficientNet, MobileNet, DenseNet—were evaluated using transfer ImageNet. achieved highest test set 96%, outperforming ResNet (91%), EfficientNet MobileNet (93%). Therefore, we focused on improving performance DenseNet, while considering it as base model. To model, implemented fine-tuning approach regularization techniques, including data augmentation, dropout, batch normalization, average pooling, coupled hyperparameter optimization. enhanced model an 97.1%. Our findings demonstrate effectiveness highlighting its improve diagnostic reliability in settings.

Language: Английский

---

Automated Classification of Cell Level of HEp-2 Microscopic Images Using Deep Convolutional Neural Networks-Based Diameter Distance Features

JUCS - Journal of Universal Computer Science, Journal Year: 2023, Volume and Issue: 29(5), P. 432 - 445

Published: May 25, 2023

Abstract: To identify autoimmune diseases in humans, analysis of HEp-2 staining patterns at cell level is the gold standard for clinical practice research communities. An automated procedure a complicated task due to variations densities, sizes, shapes and patterns, overfitting features, large-scale data volume, stained cells poor quality images. Several machine learning methods that analyse classify microscope images currently exist. However, accuracy still not required medical applications computer aided diagnosis those challenges. The purpose this work automate classification from microscopic improve diagnosis. This proposes Deep Convolutional Neural Networks (DCNNs) technique into six classes based on employing level-set method via edge detection segment shape. DCNNs are designed cell-shape fundamental distance features related with types. paper investigated effectiveness our proposed over benchmarked dataset. result shows highly superior comparing other dataset state-of-the-art methods. demonstrates has an excellent adaptability across under different lab environments. accurate pattern helps increasing process future.

Language: Английский

---

CNN-Transformer Architecture Solution for Compound Facial Expression Recognition

2021 7th International Conference on Computer and Communications (ICCC), Journal Year: 2023, Volume and Issue: unknown, P. 1804 - 1808

Published: Dec. 8, 2023

True emotions can be indicated by the human facial expression of emotions. Facial recognition has vast applications in healthcare, security, artificial intelligence, e-learning, sports, agriculture and various other fields. Although significant research been conducted on basic emotions, there is currently a surge interest recognizing compound expressions field image processing. In this paper, we present method that employs vision transformer (ViT) utilizes DenseNet-121 as backbone for CFEE RAFDB datasets. The proposed outperformed improved accuracy compared to state-of-the-art (SOTA) models. obtained results demonstrate rates 66.4% dataset 72.05% dataset. was enhanced around 9.05% approximately 3.62% comparison current state art (SOTA), thanks methodology. This approach tackles task paves way future investigations detecting complex using ViT

Language: Английский

---