Convolutional Neural Network Using Transfer Learning for Brain Tumor Classification

• Published in: 2024 IEEE International Conference on Information Technology, Electronics and Intelligent Communication Systems (ICITEICS) pp. 1 - 6
• Main Authors: Sakthi, U., Malik, Kanishk
• Format: Conference Proceeding
• Language: English
• Published: IEEE 28-06-2024
• Subjects: Accuracy; Brain Tumor Classification; Convolutional Neural Network; Image segmentation; Manuals; Medical Resonance Images; Sensitivity; Streaming media; Transfer learning; Visualization
• DOI: 10.1109/ICITEICS61368.2024.10625408

Analyzing brain tumors without using human intervention is seen as a crucial field of study. Nevertheless, convolutional neural networks (CNN) can be used to do this. They have achieved remarkable success in resolving issues related to computer vision as well as numerous other issues like visual object detection, recognition, and segmentation. By using segmentation techniques that are extremely resilient to noise and cluster size sensitivity issues for automatic Region Of Interest (ROI) recognition, brain pictures are optimized for the purpose of detecting brain tumors. The great accuracy of CNNs and the lack of need for human feature extraction make them a popular choice. It is difficult to identify a brain tumor and to precisely determine its type. The performance of CNN is better than others because it is widely used in image recognition. One of the most significant and difficult problems in medical imaging is brain tumor segmentation since manual classification with human assistance might result in incorrect diagnosis and prognoses. Furthermore, handling a big volume of data makes it a challenging process. Brain tumors are very different in appearance, and the tumor and normal tissues are similar, making it difficult to distinguish tumor areas from images.