Identification and Comparison of Spine Tumor MRI Images with Segmentation Methods

Abstract

Objective: Revealing and identifying a spine tumor using the finest segmentation technique is a stretched procedure. The Image processing techniques can be done not only for tumor, but also to categorise any fragment of human body. Spine tumors are hardly developed by spine itself, but most habitually the tumors in spine are spread by the other pretentious tumor organs.  Magnetic Resonance imaging (MRI) customs tough radio waves to yield detailed images of the inside body specifically tumors.

Methods: In this work, noise deduction is processed and quite a lot of soft computing approaches of segmentation techniques are attained. The methods are compared for further processing of the image.

Result: The results give closely 99% accuracy and specificity using Otsu and Fuzzy entropywith active contour in detecting the tumor image.

Conclusion: This scrutiny benefits in identifying the segmentation technique for further feature extraction of spine tumor image and classification.