A scrutiny on brain tumour classification using deep learning technique

Abstract People of all ages with brain tumour are one of the most severe cancers, and their grade identification is a tough issue for health surveillance and automatic diagnosis by radiologists. Several approaches from deep learning had recently introduced in the Brain tumor Classification (BTC) literature to guide radiologists in enhancing analysis of research. Our findings cover the key stages of BTC approaches focused on deep learning, like pre-processing, extraction of functions, and identification, along with their contributions and restrictions. This paper further discusses BTC's convolution neural network frames by undertaking detailed trials leveraging transfer learning though without extension of records. We introduce an extensive analysis for surveys conducted as yet and latest deep learning approaches of BTC in this analysis. In addition, this review discusses relevant standard data sets used throughout BTC assessment. Finally, this review not only looks at the previous literature issues, but still it takes some efforts to explore the future in this field it also lists some tracks for research which can be taken in the upcoming reference, particularly towards customize with intelligent health care.