Segmentation of Brain Tumor using Multiple Threshold Technique

Abstract

Radiology use medical imaging techniques to comprehend the structure and physiological functions of the body in both
healthy and diseased subjects. A non-invasive method for viewing internal body structures can be performed by using magnetic
resonance imaging (MRI). MRI characterizes soft tissue more accurately than other imaging methods like CT. In the current
study, space-occupying lesions are visualized using MRI imaging. Slices of MRI data are used to analyze lesions. Single slice
analysis is inappropriate to determine the lesion’s size and volume. Hence, the MRI sequence is used to segment the lesions.
Following segmentation, we view the MRI 2D image in 3D to look for lesions, or aberrant tissue, in the brain. The lesion is
then visualized by performing clipping. This research suggests segmenting brain tumors automatically and even provides a
3D visualization for a more thorough study. Here, a space-occupying lesion is segmented from a T2 weighted Flair sequence
of MR images in DICOM format, and by employing the segmented volume, 3D rendering and clipping are made possible