Welcome to the InfoSci Platform
Medical Image Thresholding Using Genetic Algorithm and Fuzzy Membership Functions: A Comparative Study

Medical Image Thresholding Using Genetic Algorithm and Fuzzy Membership Functions: A Comparative Study

Shashwati Mishra, Mrutyunjaya Panda
Copyright: © 2019 |Volume: 8 |Issue: 4 |Pages: 21
ISSN: 2156-177X|EISSN: 2156-1761|EISBN13: 9781522567714|DOI: 10.4018/IJFSA.2019100103
Cite Article Cite Article

MLA

Mishra, Shashwati, and Mrutyunjaya Panda. "Medical Image Thresholding Using Genetic Algorithm and Fuzzy Membership Functions: A Comparative Study." IJFSA vol.8, no.4 2019: pp.39-59. http://doi.org/10.4018/IJFSA.2019100103

APA

Mishra, S. & Panda, M. (2019). Medical Image Thresholding Using Genetic Algorithm and Fuzzy Membership Functions: A Comparative Study. International Journal of Fuzzy System Applications (IJFSA), 8(4), 39-59. http://doi.org/10.4018/IJFSA.2019100103

Chicago

Mishra, Shashwati, and Mrutyunjaya Panda. "Medical Image Thresholding Using Genetic Algorithm and Fuzzy Membership Functions: A Comparative Study," International Journal of Fuzzy System Applications (IJFSA) 8, no.4: 39-59. http://doi.org/10.4018/IJFSA.2019100103

Export Reference

Mendeley
Favorite Full-Issue Download

Abstract

Thresholding is one of the important steps in image analysis process and used extensively in different image processing techniques. Medical image segmentation plays a very important role in surgery planning, identification of tumours, diagnosis of organs, etc. In this article, a novel approach for medical image segmentation is proposed using a hybrid technique of genetic algorithm and fuzzy logic. Fuzzy logic can handle uncertain and imprecise information. Genetic algorithms help in global optimization, gives good results in noisy environments and supports multi-objective optimization. Gaussian, trapezoidal and triangular membership functions are used separately for calculating the entropy and finding the fitness value. CPU time, Root Mean Square Error, sensitivity, specificity, and accuracy are calculated using the three membership functions separately at threshold levels 2, 3, 4, 5, 7 and 9. MRI images are considered for applying the proposed method and the results are analysed. The experimental results obtained prove the effectiveness and efficiency of the proposed method.

Request Access

You do not own this content. Please login to recommend this title to your institution's librarian or purchase it from the IGI Global bookstore.