There are two types of brain tumors: benign and malignant. Brain tumors can be caused by an uncontrollably high proliferation of aberrant cells in brain tissue. While a malignant brain tumor can impact nearby brain tissues and cause a person's death, a benign brain tumor does not affect nearby normal or healthy tissue. It may be necessary to discover brain tumors early to preserve patients' lives. The segmentation, detection, and extraction of an infected tumor region from Magnetic Resonance Images (MRI) is a labor-intensive and essential task performed by medical professionals. The accuracy of this task depends solely on experience, so it is imperative to address it with Computer-Aided Technology (CAD). It uses deep learning techniques such as VGG-19 to classify brain tumors to lessen this issue. The findings indicate that while 356 dense layers perform better with enhanced data, ResNet with 64 dense layers performs better overall in terms of accuracy

Brain Tumor, Deep Learning Models, Image Classification etc

Deep learning has demonstrated significant potential in enhancing brain tumor diagnosis, surpassing conventional methods in both accuracy and efficiency (Majib et al., 2021). The study's results demonstrate that the ResNet deep learning model attained superior accuracy in brain tumor detection relative to classic VGG and Inception approaches. This indicates that progress in deep learning for medical image analysis could significantly enhance the precision of tumor detection. The research underscores the significance of choosing the suitable deep learning architecture and evaluation measures for evaluating model performance. The findings underscore the necessity for ongoing research and development in deep learning to enhance tumor detection in medical imaging. According to this performance analysis, ResNet has superior accuracy relative to other models. Subsequently, we chose ResNet for additional examination. According to this distinct layer, ResNet with 64 dense layers demonstrates superior performance in accuracy without augmentation, whereas 356 dense layers exhibit enhanced accuracy with augmented data

Pawan Kumar and Er. Tarun Dhiman (2024). Tumor Detection and Classification of MRI Brain Image using Transfer Learning Model. International Journal on Emerging Technologies, 15(2): 68–72