Brain Tumor Detection Using Image Processing

Tumors are excess cells that grow in an uncontrolled way and are one of the most common causes of cancer. Among the various types of tumors, brain tumors are more dangerous and needs immediate medical attention. Hence there is a need for fast, automated, efficient and reliable technology for accurate detection of tumors. Usually, Doctors look at MRI images of the patient's brain to locate brain tumors manually. This leads to an inaccurate detection of tumors and is very time consuming. Various image processing techniques and the advancements in artificial intelligence have made the automatic detection of brain tumors easier. In the proposed work the deep learning architecture such as VGG 19, Resnet 50 and EfficientNetB0 are used to recognize and detect the brain tumor. A dataset consisting of human brain images with tumor and without tumors which are provided as training, testing and validation data. Based on the obtained results in our proposed work, the architecture EfficientNetB0 provides a value of 96.50% accuracy and a value of 0.021 loss when compared to other three models.