Year: 2024 | Month: May | Volume: 14 | Issue: 5 | Pages: 176-180
DOI: https://doi.org/10.52403/ijhsr.20240520
Brain Tumor Classification Using Pretained Deep Convolutional Neural Networks
M. Meena1, U. Balaswetha2, M. Harini3, N. Harini4, S. Mathumitha5
1Assistant Professor/Information Technology, 2Student/Information Technology, 3Student/Information Technology, 4Student/Information Technology, 5Student/Information Technology
Vivekanandha College of Technology for Women, Anna University, Namakkal, India.
Corresponding Author: M. Meena., AP/IT
ABSTRACT
Due to their complexity and sensitivity, classifying brain diseases is a very difficult task. Because brain tumors are serious and sometimes fatal, early detection and diagnosis are essential for developing an efficient treatment plan. A vital medical imaging tool, magnetic resonance imaging (MRI) allows for the detailed, non-invasive visualization of the internal structures of the brain. When it comes to diagnosing and treating brain tumors, magnetic resonance imaging (MRI) plays a critical role. Starting with dataset preprocessing, the method applies to MRI scans and clinical data from people with different brain conditions, including cases of tumors and non-tumors. Training and testing sets make up the dataset. MRI tumor detection requires a number of processes, including feature extraction, classification, and image post-processing. For classifying brain images, the system makes use of Convolutional Neural Networks with Long Short-Term memory (LSTM) a pre-trained model using the approach of transfer learning. The proposed framework not only uses the pre-trained model to improve the performance of training a better model but also uses thresholding to improve the dataset for better accuracy and data augmentation for increasing the number of images in the dataset. Preliminary outcome shows that the family of models of Hybrid algorithm performs better than previous CNN architectures because to scale all dimensions of depth, width, and resolution of an image with a constant ratio it uses the compound coefficient. The results also demonstrated that by scaling the baseline architecture the model is able to capture complicated features and thus the overall performance of the model is improved.
Key words: Brain tumor classification, convolutional neural network, medical imaging, deep learning, transfer learning.