UPM Institutional Repository

Evaluation of performance for different filtering methods in CT brain images


Citation

Mamat, Nurwahidah and Wan Abdul Rahman, Wan Eny Zarina and Cik Soh, Shaharuddin and Mahmud, Rozi (2018) Evaluation of performance for different filtering methods in CT brain images. AIP Conference Proceedings, 2016 (1). pp. 1-9. ISSN 0094-243X; ESSN: 1551-7616

Abstract

This paper presents the comparison of filtering methods for a contrast enhancement of computed tomography (CT) brain images. Each method consists of three filter consecutively which is a combination of the low order linear filter such as Gaussian filter, disk filter, average filter and median filter with an adaptive filter method and unsharp filter. The process starts with filtering the CT brain image using low order linear filter, then proceeds with adaptive averaging filter and ends with unsharp filter. In this paper, there are two criteria, peak signal to noise ratio and mean square error, that were adopted for performance assessment. Our preliminary results showed that the combination of Gaussian filter with adaptive filter and unsharp filter gives the good result in removing the noise and edge detection. This method improved the CT brain image and the gyri and sulci can be easily identified.


Download File

[img] Text
Evaluation of performance for different filtering methods in CT brain images.pdf

Download (49kB)
Official URL or Download Paper: https://aip.scitation.org/doi/10.1063/1.5055479

Additional Metadata

Item Type: Article
Divisions: Faculty of Medicine and Health Science
DOI Number: https://doi.org/10.1063/1.5055479
Publisher: American Institute of Physics
Keywords: CT brain images; computed tomography brain images; Filtering method
Depositing User: Nurul Ainie Mokhtar
Date Deposited: 18 Nov 2020 15:19
Last Modified: 18 Nov 2020 15:19
Altmetrics: http://www.altmetric.com/details.php?domain=psasir.upm.edu.my&doi=10.1063/1.5055479
URI: http://psasir.upm.edu.my/id/eprint/72640
Statistic Details: View Download Statistic

Actions (login required)

View Item View Item