Review on Brain Tumor Detection and Classification with MRI Images using Deep Learning Techniques

  • Rashmi A. Jolhe et al.


Segmentation allows visualizing size and position of tumor within the brain which is used
for comparison of tumor images before and after surgery. Clinical Experts performs this
task manually. As clinical data is increasing day by day, the manual segmentation methods
may not be always accurate and may lead to errors for different types of brain tumors. The
image data has various attributes which are case specific and the distinction and
interdependence of these attributes plays a significant role in classification and detection.
Analyzing the data for its better understanding as per the required format finally allows us
to better detect and classify the tumor. As there is variations in the appearance of tumor
tissues among different patients, manual Segmentation is a challenging process,.
Researchers are working in this area for many years and they has presented the results
which are found to be useful. However, there is a scope for further improvement in the area
of classification and detection with respect to various quality parameters associated with
the process. Inthe clinical practice, accuracy of tumor detection is highly dependent on the
operator’s experience. In the present study, Magnetic Resonance Imaging (MRI) images
of different types of brain tumors shall be considered. Automatic segmentation and
classification of brain tumor using Deep neural network is proposed in this study. The
network shall be trained on publicly available fig-share brain tumor dataset also the MRI
images collected from Tata Memorial Hospital, Mumbai.