NeuroSight: A Deep‐Learning Integrated Efficient Approach to Brain Tumor Detection

Engineering Reports, Год журнала: 2025, Номер 7(1)

Опубликована: Янв. 1, 2025

ABSTRACT Brain tumors pose a significant health risk and require immediate attention. Despite progress, accurately classifying these remains challenging due to their location, shape, size variability. This has led exploring deep learning machine in biomedical imaging, particularly processing analyzing Magnetic Resonance Imaging (MRI) data. study compared newly developed Convolutional Neural Network model pre‐trained models using transfer learning, focusing on comprehensive comparison involving VGG‐16, ResNet‐50, AlexNet, Inception‐v3. VGG‐16 outperformed all other with 95.52% test accuracy, 99.87% training 0.2348 validation loss. ResNet‐50 got 93.31% 98.78% 0.6327 The CNN 0.2960 loss, 92.59% 98.11% accuracy. worst seemed be Inception‐v3, 89.40% 97.89% 0.4418 approach facilitates deep‐learning researchers identifying categorizing brain cancers by comparing recent papers assessing methodologies.

Язык: Английский

---

Enhanced Fox Optimizer for Internet of Things Powered Personalized Healthcare Systems

International Journal of Communication Systems, Год журнала: 2025, Номер 38(7)

Опубликована: Март 31, 2025

ABSTRACT The Internet of Things (IoT) paradigm has recently opened up new research opportunities in many academic and industrial fields, particularly medicine. IoT‐enabled technology transformed healthcare from a centralized model to personalized system driven by ubiquitous wearable devices smartphones. implementation IoT faces critical challenges, including energy efficiency, network reliability, task response time, availability services. An Adaptive Fox Optimizer (AFO) is proposed as novel IoT‐supported method for providing zero‐orientation nature AFO mitigated quasi‐oppositional learning. A reinitialization plan also presented improve exploration skills. Furthermore, an additional stage implemented with two movement techniques optimize search capabilities. In addition, multi‐best methodology used deviate the local optimum manage population more efficiently. Ultimately, greedy selection accelerates convergence exploitability. was rigorously evaluated, demonstrating significant improvements across key performance metrics. Compared conventional approaches, enhances 83.33%, reliability 11.32%, reduces consumption 19.12%, decreases times 25.14%. These results highlight AFO's ability resource allocation, enhance fault tolerance, prolong lifespan environments. By addressing this contributes developing efficient, reliable, responsive systems, paving way advancements health monitoring, telemedicine, smart hospital management.

Язык: Английский

---

Hybrid Ensemble Deep Learning Model for Advancing Ischemic Brain Stroke Detection and Classification in Clinical Application

Journal of Imaging, Год журнала: 2024, Номер 10(7), С. 160 - 160

Опубликована: Июль 2, 2024

Ischemic brain strokes are severe medical conditions that occur due to blockages in the brain’s blood flow, often caused by clots or artery blockages. Early detection is crucial for effective treatment. This study aims improve and classification of ischemic clinical settings introducing a new approach integrates stroke precision enhancement, ensemble deep learning, intelligent lesion segmentation models. The proposed hybrid model was trained tested using dataset 10,000 computed tomography scans. A 25-fold cross-validation technique employed, while model’s performance evaluated accuracy, precision, recall, F1 score. findings indicate significant improvements accuracy different stages images when enhanced SPEM with contrast-limited adaptive histogram equalization set 4. Specifically, showed improvement (from 0.876 0.933) hyper-acute images; from 0.881 0.948 acute images, 0.927 0.974 sub-acute 0.928 0.982 chronic images. Thus, shows promise strokes. Further research needed validate its on larger datasets enhance integration into settings.

Язык: Английский

---

Enhancing Acute Ischemic Stroke Diagnosis Using IoMT and Deep Learning Technologies

2022 International Telecommunications Conference (ITC-Egypt), Год журнала: 2024, Номер unknown, С. 313 - 319

Опубликована: Июль 22, 2024

Язык: Английский

---