Gliomas represent the most common form of brain tumors. The most often used technique, to establish the diagnosis, is based on magnetic resonance imaging. To establish the diagnosis in the early stage is sometimes very difficult even for a specialist with much experience, thus an efficient and reliable system is needed that helps the specialist in the interpretation. The convolutional neural network has excellent achievement in image classification; though, adjusting the values of hyperparameters is a very time-consuming process. In this paper, we propose to adjust the hyperparameters of convolutional neural networks by a modified firefly algorithm and apply it to glioma grade classification. We evaluated the proposed approach on magnetic resonance images from more data collections. The typical brain images are obtained from the IXI dataset. The glioma brain tumor images are used from the cancer imaging archive. The obtained results confirm superiority related to other techniques in the same research area.
CITATION STYLE
Bezdan, T., Zivkovic, M., Tuba, E., Strumberger, I., Bacanin, N., & Tuba, M. (2021). Glioma Brain Tumor Grade Classification from MRI Using Convolutional Neural Networks Designed by Modified FA. In Advances in Intelligent Systems and Computing (Vol. 1197 AISC, pp. 955–963). Springer. https://doi.org/10.1007/978-3-030-51156-2_111
Mendeley helps you to discover research relevant for your work.