Multiclass Brain Tumor Detection using Deep Transfer Learning

• Published in: 2024 MIT Art, Design and Technology School of Computing International Conference (MITADTSoCiCon) pp. 1 - 5
• Main Authors: Deshmukh, Aishwarya V., Bendre, M.R.
• Format: Conference Proceeding
• Language: English
• Published: IEEE 25-04-2024
• Subjects: Accuracy; brain tumour; CNN; Deep Transfer Learning; etc; Feature extraction; Image analysis; Image segmentation; Inception-V3 algorithm; Magnetic resonance imaging; Real-time systems; ResNet-50 method; Transfer learning; VGG16
• DOI: 10.1109/MITADTSoCiCon60330.2024.10575380

Brain tumors, identified as abnormal growth of cell inside the brain, necessitate accurate categorization for tailored treatment plans. This classification aids in determining the severity of the condition and guiding therapeutic decisions. Iimaging technologies in medical like Magnetic Resonance Imaging (MRI) are instrumental in identifying brain pathologies without emitting ionizing radiation. Deep learning, a subset of AI, enhances brain tumor detection from MRI scans, improving prediction rates. Digital image processing is vital for medical image analysis and diagnosis. Brain tumors manifest in various forms, categorized as benign or malignant based on their tissue characteristics. Convolutional Neural Networks (CNNs), particularly VGG-16, Inception-v3, and ResNet-50 are prominent in brain tumor analysis, segmentation, and classification. These models are compared for their efficacy in identifying and forecasting brain tumor cells. Common brain tumor types include glioma, meningioma, and pituitary tumors, each requiring specific treatment approaches.