MRI Image Segmentation Using Shannon and Non Shannon Entropy Measures

Varshika Pandey, Vipin Gupta

ABSTRACT Image Segmentation is a vital tool in medical field that enable professionals to detect their patientâ€™s problems and help them to get proper diagnosed. In this article, an entropy based approach for image segmentation is discussed to highlight tumour in MRI images. Magnetic Resonance Imaging (MRI) is a process, in which pixel values are based on radiation absorption. In the proposed approach we have selected threshold values on the basis of different entropy measures such as Shannon, Renyi, Harvard Charvart , Kapur and Vajda entropy measures to segmentize an MRI image indicating tumour. The gray level co-occurrence and probability matrix are utilized as basis functions for proposed methodology. Simulation results for different entropy measures depicts that Non Shannon Entropy measures give more promising results as compared to classical Shannon based approach, thus can be used to detect human body tumours using MRI images. Keywords: Image Segmentation, Co-occurrence and Probability matrix, Shannon Entropy, Renyi Entropy, Harvard Charvart Entropy, Kapur Entropy, and Vajda Entropy.

Title:	MRI Image Segmentation Using Shannon and Non Shannon Entropy Measures
Author Name:	Varshika Pandey, Vipin Gupta
Abstract:	ABSTRACT Image Segmentation is a vital tool in medical field that enable professionals to detect their patientâ€™s problems and help them to get proper diagnosed. In this article, an entropy based approach for image segmentation is discussed to highlight tumour in MRI images. Magnetic Resonance Imaging (MRI) is a process, in which pixel values are based on radiation absorption. In the proposed approach we have selected threshold values on the basis of different entropy measures such as Shannon, Renyi, Harvard Charvart , Kapur and Vajda entropy measures to segmentize an MRI image indicating tumour. The gray level co-occurrence and probability matrix are utilized as basis functions for proposed methodology. Simulation results for different entropy measures depicts that Non Shannon Entropy measures give more promising results as compared to classical Shannon based approach, thus can be used to detect human body tumours using MRI images. Keywords: Image Segmentation, Co-occurrence and Probability matrix, Shannon Entropy, Renyi Entropy, Harvard Charvart Entropy, Kapur Entropy, and Vajda Entropy.
Cite this article:	Varshika Pandey, Vipin Gupta , " MRI Image Segmentation Using Shannon and Non Shannon Entropy Measures " , International Journal of Application or Innovation in Engineering & Management (IJAIEM), Volume 3, Issue 7, July 2014 , pp. 041-046 , ISSN 2319 - 4847.
Full Text [PDF] Home