SAlexNet: Superimposed AlexNet using Residual Attention Mechanism for Accurate and Efficient Automatic Primary Brain Tumor Detection and Classification

Results in Engineering, Journal Year: 2025, Volume and Issue: unknown, P. 104025 - 104025

Published: Jan. 1, 2025

Language: Английский

---

ACU-Net: Attention-based convolutional U-Net model for segmenting brain tumors in fMRI images

Digital Health, Journal Year: 2025, Volume and Issue: 11

Published: Jan. 1, 2025

Objective Accurate segmentation of brain tumors in medical imaging is essential for diagnosis and treatment planning. Current techniques often struggle with capturing complex tumor features are computationally demanding, limiting their clinical application. This study introduces the attention-based convolutional U-Net (ACU-Net) model, designed to improve accuracy efficiency fMRI images by incorporating attention mechanisms that selectively highlight critical while preserving spatial context. Methods The ACU-Net model combines neural networks (CNNs) enhance feature extraction coherence. We evaluated on BraTS 2018 2020 datasets using rigorous data splitting training, validation, testing. Performance metrics, particularly Dice scores, were used assess across different regions, including whole (WT), core (TC), enhancing (ET) classes. Results demonstrated high accuracy, achieving scores 99.23%, 99.27%, 96.99% WT, TC, ET, respectively, dataset, 98.72%, 98.40%, 97.66% ET dataset. These results indicate effectively captures boundaries subregions precision, surpassing traditional approaches. Conclusion shows significant potential planning providing precise efficient images. integration within a CNN architecture proves beneficial identifying structures, suggesting can be valuable tool applications.

Language: Английский

---

ViT-CB: Integrating hybrid Vision Transformer and CatBoost to enhanced brain tumor detection with SHAP

Biomedical Signal Processing and Control, Journal Year: 2024, Volume and Issue: 100, P. 107027 - 107027

Published: Oct. 24, 2024

Language: Английский

---

A multi-stage fusion deep learning framework merging local patterns with attention-driven contextual dependencies for cancer detection

Computers in Biology and Medicine, Journal Year: 2025, Volume and Issue: 189, P. 109916 - 109916

Published: March 6, 2025

Language: Английский

---

Deep Learning Transformations in Medical Imaging: Advancements in Brain Tumor, Retinal Vessel, and Inner Ear Segmentation

Published: Jan. 2, 2025

Traditional approaches to segmenting retinal vessels, brain tumors, and inner ear structures are typically based on extracting features manually. Machine learning such as support vector machines (SVMs) random forests commonly used for these domains. Although methods have generated encouraging results, they frequently limited by the handcrafted feature quality sensitivity of machine algorithms hyperparameters. Deep has sparked a revolution in segmentation tasks automating extraction complex from medical images, eliminating need manual engineering. As result, imaging techniques improving terms accuracy robustness. In realm imaging, deep primarily leverage convolutional neural network (CNN) architectures. Notably, models U-Net, Attention ResUNet gained popularity 3D image segmentation. CNNs excel at producing probability maps each pixel input image, indicating likelihood belonging specific class. However, fall short capturing channel-level information. To address this limitation, newer variations squeeze excitation embedded attention U-Net (SEEA-U-Net) emerged, focusing within regions interest analysis. These advancements find extensive application tumor offer adaptability through architectural adjustments, striking fine balance between computational efficiency precision. The combination diverse CNN designs proven be robust strategy adapting various modalities. Inner structure segmentation, including cochlea preoperative CT helps diagnosis treatment hearing impairments, especially cochlear implant surgery. Experiments demonstrated exceptional performance super-resolution architecture known SRSegN (i.e. network). With mean Dice score exceeding 0.80, outperforms several established models. This builds upon encoder–decoder incorporating branch up-sampling that incrementally extracts low-resolution inputs upscales maps. chapter, we elucidate profound impact head anatomies, spanning applications detection, vessel self-learning capabilities inherent ushered transformative era, significantly enhancing effectiveness outcomes critical tasks.

Language: Английский

---

MRI intracranial Neoplasm classification using hybrid LOA-based deep learning classifier

Biomedical Signal Processing and Control, Journal Year: 2025, Volume and Issue: 104, P. 107560 - 107560

Published: Jan. 28, 2025

Language: Английский

---

Advanced Brain Tumor MR Image Classification Using a Combination Undecimated Wavelet Transform, EfficientNet-B0 and PCA via Multi-SVM Analysis

Iranian Journal of Science and Technology Transactions of Electrical Engineering, Journal Year: 2025, Volume and Issue: unknown

Published: March 12, 2025

Language: Английский

---

EC-HDLNet: Extended coati-based hybrid deep dilated convolutional learning network for brain tumor classification

Biomedical Signal Processing and Control, Journal Year: 2025, Volume and Issue: 107, P. 107865 - 107865

Published: March 26, 2025

Language: Английский

---

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

Discover Artificial Intelligence, Journal Year: 2025, Volume and Issue: 5(1)

Published: March 27, 2025

Language: Английский

---

The Role of Artificial Intelligence in Diagnostic Neurosurgery: A Systematic Review

Research Square (Research Square), Journal Year: 2025, Volume and Issue: unknown

Published: April 4, 2025

Background: Artificial intelligence (AI) is increasingly applied in diagnostic neurosurgery, enhancing precision and decision-making neuro-oncology, vascular, functional, spinal subspecialties. Despite its potential, variability outcomes necessitates a systematic review of performance applicability. Methods: A comprehensive search PubMed, Cochrane Library, Embase, CNKI, ClinicalTrials.gov was conducted from January 2020 to 2025. Inclusion criteria comprised studies utilizing AI for reporting quantitative metrics. Studies were excluded if they focused on non-human subjects, lacked clear metrics, or did not directly relate applications neurosurgery. Risk bias assessed using the PROBAST tool. This study registered PROSPERO, number CRD42025631040 26^th, Results: Within 186 studies, neural networks (29%) hybrid models (49%) dominated. categorised into neuro-oncology (52.69%), vascular neurosurgery (19.89%), functional (16.67%), (11.83%). Median accuracies exceeded 85% most categories, with achieving high accuracy tumour detection, grading, segmentation. Vascular excelled stroke intracranial haemorrhage median AUC values 97%. Functional showed promising results, though sensitivity specificity underscores need standardised datasets validation. Discussion: The review’s limitations include lack data weighting, absence meta-analysis, limited collection timeframe, quality, risk some studies. Conclusion: AI shows potential improving across neurosurgical domains. Models used stroke, ICH, aneurysm conditions such as Parkinson’s disease epilepsy demonstrate results. However, sensitivity, specificity, further research model refinement ensure clinical viability effectiveness.

Language: Английский

---