Abstract Deep Learning Algorithms for medical image analysis is increasing day by day, particularly in Radiology. Tumors in many parts of the body are malignant/non-malignant and should be identified as early as possible. Due to the complex structure of the brain and the existence of more noise in the scanned images, manual identification of tumors in the brain becomes harder for health care experts and it is time consuming. Hence in this proposed work, Visual Geometry Group (VGG) a classical convolution neural network(CNN) is developed in oncology to solve the problem of early identification and detection. CNN is the most effective technique for the classification of tumor and non-tumor tissues in early stage which embrace pre-processing of image followed by feature extraction, and succeeding classification. The proposed model uses VGG 16 which consists of 16 convolution layers that classifies images into 1000 different categories. It is trained and tested by using images in the BRATS dataset that shows the accuracy of about 98.75% compared to the state of art methods.
Alan : Mühendislik
Dergi Türü : Uluslararası
Benzer Makaleler | Yazar | # |
---|
Makale | Yazar | # |
---|