An Automatic Detection of Brain Tumor using CNN & VGG19

Abstract

According to the 2019 cancer statistics by WHO, brain tumors are considered the main cause of mortality related to cancer throughout the world and are known as one of the most common forms of cancer both in children as well as adults. Among the most common brain tumors, we have those that begin and tend to remain in the brain, which as meningioma with 34% of presence, another type of tumor is called glioma, arising from the surrounding tissue in the brain, it is part of 30% of all tumors in the brain, however, this glioma represents 80% of malignant tumors, making it the most common tumor common that causes death. However, this scheme depicts how convolutional neural networks using VGG19 model can provide an effective mechanism to detect brain tumors at an early stage using MRI images and can save the lives of mankind. Consequently, this research classified glioma brain tumor images using VGG-19 with HE preprocessing data. The model was tested to get a comparison of accuracy, precision, recall, and f1-score of the two test data, namely the original data and HE data. Based on the results of model testing, we can see in table2 that the original data produced the highest values of accuracy, precision, recall, and F1-score, with values of 97% accuracy, 100% precision, 97% recall, and 98% f1 score. While data using HE preprocessing has an accuracy value of 92%, precision of 100%, recall of 92%, and f1 score of 96%t..