Detection and 3D Visualization of Brain Tumor using Deep Learning and Polynomial Interpolation

Among different imaging techniques MRI, MRSI and CT scans are some of the widely use techniques to visualize brain structures to point out brain anomalies especially brain tumor. Identification of brain tumor accurately in clinical practices has always been a hard decision for neurologist as multiple exceptions might present in images which may lead dubious suggestion from neurologist. In our proposed model we are aiming towards brain tumor detection and 3d visualization of tumor more accurately in effcient way. Our proposed model composed of three stages such as classification of image using CNN whether any tumor exists of not; segmentation using multi-thresholding to extract the detected tumor; and 3d visualization using polynomial interpolation. the proposed model enables enhancing the accuracy of tumor detection as compare to existing models as well as segmenting and 3d visualizing the detected tumor. we get 85% accuracy on our model comparing with others which is slightly more efficient in terms of classification and detection.

I. Introduction

Magnetic resonance imaging (MRI) has become essential part of medical science. It has several advantages over other medical imaging modalities with many applications, such as, cardiovascular, neurological, and in particular brain imaging. From last decade MRI images are become more popular to identify the tumors from brain [1].