|
Effective Detection of Brain Tumor from Colour MRI Images Using Clustering Segmentation Techniques
|
|
| Author:
|
S SANTHIKUMARI, VAMSIDHAR ENIREDDY
|
| Abstract:
|
Brain is the primary complex organ in the human body. Brain tumor due to the abnormal cells of proliferation,
brain tumor decrease the life expectancy of the patient when not identify at early stage. If identify at an early
stage, life expectation of the patientmay rise. To identify the tumor preciselyso many image processing
techniques were proposed by so many authors.Nowadays proper segmentation of tumors in brain images is
very necessarythat helps in proper treatment. Among various modalities Magnetic resonance imaging (MRI)is
usedto segment the tumor due low ionization and noise. In this paper, study the various clustering
segmentation techniques to identify the tumor from human brain MRI images. The clustering techniques such
as K-means andfuzzy c-means(FCM) techniques are used after proper pre-processing steps. Both clustering
methods performances evaluated by dice coefficient (DC), Jaccard similarity index (JSI),structural similarity
indexmeasure (SSIM), and time. Finally,the FCM accurately detect the tumor as compared to the K-means
clustering due to multiple clusters update iteratively. And it is observed that FCM take more time to identify
the tumoras compared with K-means clustering algorithm.
|
| Keyword:
|
Brain tumor, MRI, K-means, fuzzy c-means,dice coefficient, SSIM, JSI
|
| EOI:
|
-
|
| DOI:
|
https://doi.org/10.31838/ijpr/2021.13.02.182
|
| Download:
|
Request For Article
|
|
|