NeuroNet19: an explainable deep neural network model for the classification of brain tumors using magnetic resonance imaging data

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

Опубликована: Янв. 17, 2024

Abstract Brain tumors (BTs) are one of the deadliest diseases that can significantly shorten a person’s life. In recent years, deep learning has become increasingly popular for detecting and classifying BTs. this paper, we propose neural network architecture called NeuroNet19. It utilizes VGG19 as its backbone incorporates novel module named Inverted Pyramid Pooling Module (iPPM). The iPPM captures multi-scale feature maps, ensuring extraction both local global image contexts. This enhances maps produced by backbone, regardless spatial positioning or size tumors. To ensure model’s transparency accountability, employ Explainable AI. Specifically, use Local Interpretable Model-Agnostic Explanations (LIME), which highlights features areas focused on while predicting individual images. NeuroNet19 is trained four classes BTs: glioma, meningioma, no tumor, pituitary tested public dataset containing 7023 Our research demonstrates achieves highest accuracy at 99.3%, with precision, recall, F1 scores 99.2% Cohen Kappa coefficient (CKC) 99%.

Язык: Английский

---

A multi-modality framework for precise brain tumor detection and multi-class classification using hybrid GAN approach

Biomedical Signal Processing and Control, Год журнала: 2025, Номер 104, С. 107559 - 107559

Опубликована: Фев. 11, 2025

Язык: Английский

---

Brain tumor classification from MRI images using exponential-Walruses hunting optimization driven SqueezeNet

Expert Systems with Applications, Год журнала: 2025, Номер 271, С. 126633 - 126633

Опубликована: Янв. 23, 2025

Язык: Английский

---

Cross prior Bayesian attention with correlated inception and residual learning for brain tumor classification using MR images (CB-CIRL Net)

Journal of Neuroscience Methods, Год журнала: 2025, Номер unknown, С. 110392 - 110392

Опубликована: Фев. 1, 2025

Язык: Английский

---

Brain tumor diagnosis using modified DenseNet121 architecture with adaptive learning rate and callback mechanism

Neural Computing and Applications, Год журнала: 2025, Номер unknown

Опубликована: Март 19, 2025

Язык: Английский

---

From black box AI to XAI in neuro-oncology: a survey on MRI-based tumor detection

Discover Artificial Intelligence, Год журнала: 2025, Номер 5(1)

Опубликована: Март 27, 2025

Язык: Английский

---

A Novel Hybrid Machine Learning-Based System Using Deep Learning Techniques and Meta-Heuristic Algorithms for Various Medical Datatypes Classification

Diagnostics, Год журнала: 2024, Номер 14(14), С. 1469 - 1469

Опубликована: Июль 9, 2024

Medicine is one of the fields where advancement computer science making significant progress. Some diseases require an immediate diagnosis in order to improve patient outcomes. The usage computers medicine improves precision and accelerates data processing diagnosis. In categorize biological images, hybrid machine learning, a combination various deep learning approaches, was utilized, meta-heuristic algorithm provided this research. addition, two different medical datasets were introduced, covering magnetic resonance imaging (MRI) brain tumors other dealing with chest X-rays (CXRs) COVID-19. These introduced network that contained techniques, which based on convolutional neural (CNN) or autoencoder, extract features combine them next step select optimal using particle swarm optimization (PSO) algorithm. This sought reduce dimensionality while maintaining original performance data. considered innovative method ensures highly accurate classification results across datasets. Several classifiers employed predict diseases. COVID-19 dataset found highest accuracy 99.76% CNN-PSO-SVM. comparison, tumor obtained 99.51% accuracy, derived autoencoder-PSO-KNN.

Язык: Английский

---

Detection of Alzheimer's disease using deep learning models: A systematic literature review

Informatics in Medicine Unlocked, Год журнала: 2024, Номер 50, С. 101551 - 101551

Опубликована: Янв. 1, 2024

Alzheimer's disease (AD) is a progressive neurological considered the most common form of late-stage dementia. Usually, AD leads to reduction in brain volume, impacting various functions. This article comprehensively analyzes context fivefold main topics. Firstly, it reviews imaging techniques used diagnosing disease. Secondly, explores proposed deep learning (DL) algorithms for detecting Thirdly, investigates commonly datasets develop DL techniques. Fourthly, we conducted systematic review and selected 45 papers published highly ranked publishers (Science Direct, IEEE, Springer, MDPI). We analyzed them thoroughly by delving into stages diagnosis emphasizing role preprocessing Lastly, paper addresses remaining practical implications challenges context. Building on analysis, this survey contributes covering several aspects related that have not been studied thoroughly.

Язык: Английский

---

A Linear time shrinking-SL(t)-ViT approach for brain tumor identification and categorization

IETE Journal of Research, Год журнала: 2024, Номер 70(11), С. 8300 - 8322

Опубликована: Авг. 28, 2024

Automated brain tumor detection and classification systems have gained popularity in recent years because traditional diagnosis procedures are time-consuming costly nature. Deep learning(DL) methods, specifically pre-trained convolutional neural networks (CNNs), shown promising results accurately rapidly classifying tumors. However, the lack of diverse magnetic resonance imaging (MRI) datasets has hindered ability DL algorithms to generalize effectively. To address this issue, paper proposes a model using generative adversarial (GANs) conjunction with data augmentation strategies structural similarity loss function employed for generating annotated images. A novel inspired from Vision Transformer, Shrinking Linear Time Transformer (SL(t)-ViT) network is proposed disease classification. The underwent extensive evaluation across multiple datasets, employing standard performance metrics assess its efficacy identification. achieved remarkable testing accuracies 0.995, 0.996, 0.9954, 0.998, 0.997 binary tasks 0.986, 0.982, 0.985, 0.993 multi-class tasks. These underscore superior our model, showcasing capability outperform state-of-the-art techniques. Specifically, it demonstrated substantial margin improvement, ranging 1-2% 9-10% classification, solidifying position as leading approach Results demonstrate outperforming other models. Overall, study highlights potential SL(t)-ViT GANs improving accuracy, resource consumption, efficiency diagnosis.

Язык: Английский

---

Deep learning and transfer learning for brain tumor detection and classification

Biology Methods and Protocols, Год журнала: 2024, Номер 9(1)

Опубликована: Янв. 1, 2024

Convolutional neural networks (CNNs) are powerful tools that can be trained on image classification tasks and share many structural functional similarities with biological visual systems mechanisms of learning. In addition to serving as a model systems, CNNs possess the convenient feature transfer learning where network one task may repurposed for training another, potentially unrelated, task. this retrospective study public domain MRI data, we investigate ability models brain cancer imaging data while introducing unique camouflage animal detection step means enhancing networks' tumor ability. Training glioma normal post-contrast T1-weighted T2-weighted, demonstrate potential success strategy improving accuracy. Qualitative metrics such space DeepDreamImage analysis internal states were also employed, which showed improved generalization by following Image saliency maps further investigation allowing us visualize most important regions from network's perspective Such methods not only 'look' at itself when deciding, but impact surrounding tissue in terms compressions midline shifts. These results suggest an approach MRIs is comparable radiologists exhibiting high sensitivity subtle changes resulting presence tumor.

Язык: Английский

---