DocumentCode :
3687273
Title :
Brain FMRI clustering using interaction K-means algorithm with PCA
Author :
K. Vijay;K. Selvakumar
Author_Institution :
Dept. of Information Technology, Faculty of Computing, Sathyabama University, Chennai-600119, Tamil Nadu, India
fYear :
2015
fDate :
4/1/2015 12:00:00 AM
Firstpage :
909
Lastpage :
913
Abstract :
An uncontrolled development of tissues in the human being is a tumor. To spot exact locality of tumor and its data being an essential task. To find the tumor, magnetic resonance image has used. FMRI (Functional Magnetic Resonance Imaging) is non-invasive procedure is to see the brain activity and to calculate of blood circulation level in the brain. The FMRI data are time series of 3D-dimensional volume images of the brain based on interaction pattern. This data is traditionally analyzed within a mass-univariate framework essentially relying on classical inferential statistics. Segmentation of FMRI plays a vital role to acquire the data of brain activities of the human brain. Feature Selection is a complicated to handle in Interaction pattern. To overcome the difficulty of the feature selection process, we use the Principal Component Analysis (PCA). PCA is a technique to pre-process the data before carrying out any data mining responsibilities, e.g., categorization and grouping. By this PC Analysis, profitable predictors are achievable in FMRI data. Using the features selected from PCA, we present a clustering novel technique. Based on this new cluster notion interaction K-means (IKM) have applied. IKM is a well-organized procedure for clustering. In this paper, Trained MRI data has taken out using the PC Analysis and IKM technique have applied over the specified data. By this, improvement has conquered in the performance in terms of accuracy and complexity in multivariate data.
Keywords :
"Indexes","Algorithm design and analysis","Imaging","Merging"
Publisher :
ieee
Conference_Titel :
Communications and Signal Processing (ICCSP), 2015 International Conference on
Type :
conf
DOI :
10.1109/ICCSP.2015.7322628
Filename :
7322628
Link To Document :
بازگشت