Deep Neural Networks for Precise Brain Tumor Delineation: A U-Net and TensorFlow Approach

Brain tumors, especially gliomas, are complex and aggressive growths of cells in the brain that lead to high morbidity and mortality. With high-grade gliomas having a median survival rate of under two years, accurate and timely diagnosis is crucial. Magnetic Resonance Imaging (MRI) is the primary me...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Engineering, technology & applied science research Ročník 15; číslo 3; s. 23686 - 23691
Hlavní autoři: Kundur, Niranjan C., Divakar, H. R., Khaiyum, Samitha, Rakshitha, Kiran P., Dhulavvagol, Praveen M., Meti, Anand S.
Médium: Journal Article
Jazyk:angličtina
Vydáno: 04.06.2025
ISSN:2241-4487, 1792-8036
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Brain tumors, especially gliomas, are complex and aggressive growths of cells in the brain that lead to high morbidity and mortality. With high-grade gliomas having a median survival rate of under two years, accurate and timely diagnosis is crucial. Magnetic Resonance Imaging (MRI) is the primary method for detecting brain tumors, but manual interpretation by radiologists can be time-consuming and subject to variability. Therefore, there is a growing need for more reliable and automated methods. This study proposes a deep learning approach for Brain Tumor Segmentation (BraTS) using the U-Net model in TensorFlow. U-Net is well-suited for biomedical image segmentation due to its encoder-decoder structure and skip connections, which capture detailed information and spatial context. The model is trained on the BraTS 2020 dataset, which includes MRI scans of high-grade and low-grade gliomas across four sequences: Fluid-Attenuated Inversion Recovery (FLAIR), T1-weighted, T1-weighted with Contrast Enhancement (T1CE), and T2-weighted. This work demonstrates the potential of deep learning to improve medical imaging precision, enhancing diagnosis and treatment planning for brain tumor patients.
ISSN:2241-4487
1792-8036
DOI:10.48084/etasr.10684