Submit Manuscript  

Article Details


Brain Tumor segmentation of T1w MRI images based on Clustering using Dimensionality Reduction Random Projection Technique

Author(s):

K. Rajesh Babu*, P. V. Nagajaneyulu and K. Satya Prasad   Pages 1 - 15 ( 15 )

Abstract:


Background: Early diagnosis of a brain tumor may increase life expectancy. If not diagnosed at an early stage, the brain tumor shortens the life expectancy of the diseased. Accompanied by several segmentation algorithms, Magnetic Resonance Imaging (MRI) preferred as a reliable assessment technique. The availability of high-dimensional medical image data during the identification procedure can place a heavy computational burden and require a suitable preprocessing step for lower-dimensional representation. At the same time, to reduce the storage requirement and complexity of the image data Random Projection Technique (RPT) is widely accepted as the multivariate approach for data reduction.

Aims: This paper mainly focuses on T1-weighted MRI images clustering for brain tumor segmentation with dimension reduction by using a conventional Principle Component Analysis (PCA) and RPT.

Methods: Two clustering algorithms, K-Means and Fuzzy C-Means, are used to detect the brain tumor. The primary objective is to present a comparison of two cluster methods between the PCA algorithm and RPT on MRIs. Apart from the original dimension of 512×512, the analysis used three other sizes, 256×256, 128×128, and 64×64, to study the effect of methods.

Results: As per the average reconstruction, Euclidean distance, and segmentation distance errors, the RPT attained better results as compared with PCA along with clustered images. According to the performance metrics, the RPT supported the Fuzzy C-Means algorithm in achieving the best clustering performance, and significant results for each resize of MRI images.

Keywords:

Dimension reduction, average reconstruction error, Euclidean distance, segmentation distance error, random projection technique, principle component analysis, fuzzy C-means, K-means.

Affiliation:

Department of Electronics and Communication Engineering, Koneru Lakshmaiah Education Foundation, Guntur, Sri Mittapalli College of Engineering, Guntur, Vignan’s Foundation for Science, Technology & Research, Guntur



Read Full-Text article