Deep Learning Model For Brain Tumor Identification (KEY IJP************815)

• Athulya A.m,Neha.s,Gayathri.n

Abstract

Developed a CNN-based model for the detection of brain tumors from MRI images which improved diagnostic accuracy and efficiency. The analysis was based on a comprehensive dataset of MRI images where the researchers applied powerful data augmentation methods and used transfer learning from ResNet architecture to make the model work optimally. The model was able to achieve a very high accuracy of 95% through fine-tuning of our model by changing the various parameters, which resulted in outperforming conventional methods. It was proven that its sensitivity and specificity were superior in the correct detection of the tumor. We concluded that the model can artfully comprehend morphing in MRI images as a key to an aloof account. The final results showed that it has the potential to provide reliable diagnosis very quickly and thus have a positive effect on a patient's health. In turn, such frameworks will thrive on real-time reporting while the dataset will become broader for generalization.