Fusion at Features Level for MRI Image Segmentation

Manoj Kumar S. Kathane; Vilas M. Thakare

Diagnostic imaging is an important tool in healthcare applications. So far, Magnetic resonance imaging (MRI), computed tomography (CT), digital mammography, and others, are providing easy means to physicians for examining patient’s condition and take decision over the particular diagnosis. Recently, MRI has become most preferred imaging modalities, especially, in brain and heart related diagnostic. Due to advances in computing hardware and its easy availability, the performance of MRI system has been improved dramatically since its inception and is able to provide fast imaging, better resolution, immunity to artifacts and cheaper cost. One of the most important problems in image processing and analysis is segmentation and same is true for biomedical imaging. Segmentation is generally a twostep process; feature extraction and classification. In this paper, we have analyzed the segmentation performance for fusion of features such as wavelet, histogram of gradients (HOG) and linear binary pattern (LBP). The classification carried out with two different classifiers; support vector machine (SVM) and neural network. the results presented in terms of precision and recall obtained in segmentation experiments for white matter and gray matter from MRI images. The result confirms the appropriateness of use of new features like HOG and LBP.

Select Volume / Issues:

• Volume 3 Issue 6, Nov. 2012

Year:

2012

Type of Publication:

Article

Keywords:

By Using Internet Integration Technique Implementation Of Embeded systems

Journal:

IJECCE

Volume:

3

Number:

6

Pages:

1298-1302

Month:

November