A novel Swin transformer approach utilizing residual multi-layer perceptron for diagnosing brain tumors in MRI images

International Journal of Machine Learning and Cybernetics, Journal Year: 2024, Volume and Issue: 15(9), P. 3579 - 3597

Published: March 5, 2024

Abstract Serious consequences due to brain tumors necessitate a timely and accurate diagnosis. However, obstacles such as suboptimal imaging quality, issues with data integrity, varying tumor types stages, potential errors in interpretation hinder the achievement of precise prompt diagnoses. The rapid identification plays pivotal role ensuring patient safety. Deep learning-based systems hold promise aiding radiologists make diagnoses swiftly accurately. In this study, we present an advanced deep learning approach based on Swin Transformer. proposed method introduces novel Hybrid Shifted Windows Multi-Head Self-Attention module (HSW-MSA) along rescaled model. This enhancement aims improve classification accuracy, reduce memory usage, simplify training complexity. Residual-based MLP (ResMLP) replaces traditional Transformer, thereby improving speed, parameter efficiency. We evaluate Proposed-Swin model publicly available MRI dataset four classes, using only test data. Model performance is enhanced through application transfer augmentation techniques for efficient robust training. achieves remarkable accuracy 99.92%, surpassing previous research models. underscores effectiveness Transformer HSW-MSA ResMLP improvements innovative diagnostic offering support diagnosis, ultimately outcomes reducing risks.

Language: Английский

---

Employing deep learning and transfer learning for accurate brain tumor detection

Scientific Reports, Journal Year: 2024, Volume and Issue: 14(1)

Published: March 27, 2024

Abstract Artificial intelligence-powered deep learning methods are being used to diagnose brain tumors with high accuracy, owing their ability process large amounts of data. Magnetic resonance imaging stands as the gold standard for tumor diagnosis using machine vision, surpassing computed tomography, ultrasound, and X-ray in its effectiveness. Despite this, remains a challenging endeavour due intricate structure brain. This study delves into potential transfer architectures elevate accuracy diagnosis. Transfer is technique that allows us repurpose pre-trained models on new tasks. can be particularly useful medical tasks, where labelled data often scarce. Four distinct were assessed this study: ResNet152, VGG19, DenseNet169, MobileNetv3. The trained validated dataset from benchmark database: Kaggle. Five-fold cross validation was adopted training testing. To enhance balance improve performance models, image enhancement techniques applied four categories: pituitary, normal, meningioma, glioma. MobileNetv3 achieved highest 99.75%, significantly outperforming other existing methods. demonstrates revolutionize field

Language: Английский

---

An automated metaheuristic-optimized approach for diagnosing and classifying brain tumors based on a convolutional neural network

Results in Engineering, Journal Year: 2024, Volume and Issue: 23, P. 102459 - 102459

Published: June 28, 2024

Brain tumors must be classified to determine their severity and appropriate therapy. Applying Artificial Intelligence medical imaging has enabled remarkable developments. The presented framework classifies patients with brain high accuracy efficiency using CNN, pre-trained models, the Manta Ray Foraging Optimization (MRFO) algorithm on X-ray MRI images. Additionally, CNN Transfer Learning (TL) hyperparameters will optimized through MRFO, resulting in improved performance of model. Two public datasets from Kaggle were used build models. first dataset consists two classes, while 2nd includes three contrast-enhanced T1-weighted classes. First, a patient should diagnosed as "Healthy" (or "Tumor"). When scan returns result "Healthy," no abnormalities brain. If reveals that tumor, an performed them. After that, type tumor (i.e., meningioma, pituitary, glioma) identified second proposed classifier. A comparative analysis models two-class showed VGG16's model outperformed other Besides, Xception achieved best results three-class dataset. manual review misclassifications was conducted reasons for correct evaluation suggested architecture yielded 99.96% X-rays 98.64% MRIs. deep learning most current

Language: Английский

---

Performance of convolutional neural networks for the classification of brain tumors using magnetic resonance imaging

Heliyon, Journal Year: 2024, Volume and Issue: 10(3), P. e25468 - e25468

Published: Feb. 1, 2024

Brain tumors are a diverse group of neoplasms that challenging to detect and classify due their varying characteristics. Deep learning techniques have proven be effective in tumor classification. However, there is lack studies compare these using common methodology. This work aims analyze the performance convolutional neural networks classification brain tumors. We propose network consisting few layers, batch normalization, max-pooling. Then, we explore recent deep architectures, such as VGG, ResNet, EfficientNet, or ConvNeXt. The study relies on two magnetic resonance imaging datasets with over 3000 images three types –gliomas, meningiomas, pituitary tumors–, well without determine optimal hyperparameters training validation sets. test sets used assess models from different perspectives, including scratch, data augmentation, transfer learning, fine-tuning. experiments performed TensorFlow Keras libraries Python. accuracy complexity based capacity networks, times, image throughput. Several achieve high rates both datasets, best model achieving 98.7% accuracy, which par state-of-the-art methods. average precision for each type 94.3% gliomas, 93.8% 97.9% tumors, 95.3% VGG largest 171 million parameters, whereas MobileNet EfficientNetB0 smallest ones 3.2 5.9 respectively. These also fastest train 23.7 25.4 seconds per epoch, On other hand, ConvNext slowest 58.2 epoch. Our custom obtained highest throughput 234.37 second, followed by 226 second. 97.35 MobileNet, EfficientNet most accurate demonstrating superior terms complexity. Most fine-tuning step. augmentation does not contribute increasing general.

Language: Английский

---

Enhancing EfficientNetv2 with global and efficient channel attention mechanisms for accurate MRI-Based brain tumor classification

Cluster Computing, Journal Year: 2024, Volume and Issue: 27(8), P. 11187 - 11212

Published: May 20, 2024

Abstract The early and accurate diagnosis of brain tumors is critical for effective treatment planning, with Magnetic Resonance Imaging (MRI) serving as a key tool in the non-invasive examination such conditions. Despite advancements Computer-Aided Diagnosis (CADx) systems powered by deep learning, challenge accurately classifying from MRI scans persists due to high variability tumor appearances subtlety early-stage manifestations. This work introduces novel adaptation EfficientNetv2 architecture, enhanced Global Attention Mechanism (GAM) Efficient Channel (ECA), aimed at overcoming these hurdles. enhancement not only amplifies model’s ability focus on salient features within complex images but also significantly improves classification accuracy tumors. Our approach distinguishes itself meticulously integrating attention mechanisms that systematically enhance feature extraction, thereby achieving superior performance detecting broad spectrum Demonstrated through extensive experiments large public dataset, our model achieves an exceptional high-test 99.76%, setting new benchmark MRI-based classification. Moreover, incorporation Grad-CAM visualization techniques sheds light decision-making process, offering transparent interpretable insights are invaluable clinical assessment. By addressing limitations inherent previous models, this study advances field medical imaging analysis highlights pivotal role enhancing interpretability learning models diagnosis. research sets stage advanced CADx systems, patient care outcomes.

Language: Английский

---

BrainNet: a fusion assisted novel optimal framework of residual blocks and stacked autoencoders for multimodal brain tumor classification

Scientific Reports, Journal Year: 2024, Volume and Issue: 14(1)

Published: March 11, 2024

Abstract A significant issue in computer-aided diagnosis (CAD) for medical applications is brain tumor classification. Radiologists could reliably detect tumors using machine learning algorithms without extensive surgery. However, a few important challenges arise, such as (i) the selection of most deep architecture classification (ii) an expert field who can assess output models. These difficulties motivate us to propose efficient and accurate system based on evolutionary optimization four types modalities (t1 tumor, t1ce t2 flair tumor) large-scale MRI database. Thus, CNN modified domain knowledge connected with algorithm select hyperparameters. In parallel, Stack Encoder–Decoder network designed ten convolutional layers. The features both models are extracted optimized improved version Grey Wolf updated criteria Jaya algorithm. speeds up process improves accuracy. Finally, selected fused novel parallel pooling approach that classified neural networks. Two datasets, BraTS2020 BraTS2021, have been employed experimental tasks obtained average accuracy 98% maximum single-classifier 99%. Comparison also conducted several classifiers, techniques, nets; proposed method achieved performance.

Language: Английский

---

Explainable artificial intelligence with UNet based segmentation and Bayesian machine learning for classification of brain tumors using MRI images

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: Jan. 3, 2025

Detecting brain tumours (BT) early improves treatment possibilities and increases patient survival rates. Magnetic resonance imaging (MRI) scanning offers more comprehensive information, such as better contrast clarity, than any alternative process. Manually separating BTs from several MRI images gathered in medical practice for cancer analysis is challenging time-consuming. Tumours scans of the are exposed utilizing methods machine learning technologies, simplifying process doctors. can sometimes appear normal even when a has tumour or malignancy. Deep approaches have recently depended on deep convolutional neural networks to analyze with promising outcomes. It supports saving lives faster rectifying some errors. With this motivation, article presents new explainable artificial intelligence semantic segmentation Bayesian tumors (XAISS-BMLBT) technique. The presented XAISS-BMLBT technique mainly concentrates classification BT images. approach initially involves bilateral filtering-based image pre-processing eliminate noise. Next, performs MEDU-Net+ define impacted regions. For feature extraction process, ResNet50 model utilized. Furthermore, regularized network (BRANN) used identify presence BTs. Finally, an improved radial movement optimization employed hyperparameter tuning BRANN To highlight performance technique, series simulations were accomplished by benchmark database. experimental validation portrayed superior accuracy value 97.75% over existing models.

Language: Английский

---

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

Biomedical Signal Processing and Control, Journal Year: 2025, Volume and Issue: 104, P. 107559 - 107559

Published: Feb. 11, 2025

Language: Английский

---

Optimized deep learning architecture for brain tumor classification using improved Hunger Games Search Algorithm

Computers in Biology and Medicine, Journal Year: 2023, Volume and Issue: 160, P. 106966 - 106966

Published: April 24, 2023

Language: Английский

---

Brain tumor segmentation and classification on MRI via deep hybrid representation learning

Expert Systems with Applications, Journal Year: 2023, Volume and Issue: 224, P. 119963 - 119963

Published: March 24, 2023

Language: Английский

---