An Enhanced Edge Detection Algorithm for Medical Image Segmentation Based on Fuzzy Logic

International Journal of Computer Science and Technology | Vol. 12, No. 11, November 2010 | pp. 76–82
DOI: 10.46882/2010/IJCST/000105

Research Article

Title: An Enhanced Edge Detection Algorithm for Medical Image Segmentation Based on Fuzzy Logic

Names of Authors: Amina T. Al-Farsi¹, Yuki Tanaka²

Authors’ Affiliations: ¹Biomedical Engineering Department, King Saud University, Riyadh, Saudi Arabia; ²Graduate School of Information Science, Kyoto University, Kyoto, Japan

Abstract: Accurate edge detection in medical images is a critical prerequisite for clinical diagnostics, tumor boundaries identification, and computer-aided surgeries. Classic differential operators such as Sobel and Canny frequently fail when applied to Magnetic Resonance Imaging (MRI) scans due to low contrast and high speckle noise levels. This paper introduces an Enhanced Fuzzy Edge Detection Algorithm (EFEDA) that incorporates adaptive membership functions to resolve pixel ambiguities. The proposed system analyzes a 3 * 3 pixel neighborhood window, computing intensity gradients which are subsequently mapped to fuzzy linguistic sets (Low, Medium, High) using Gaussian functions. Fuzzy rules determine whether a pixel belongs to a true boundary, background, or noise artifact. Tests performed on a standard dataset of 200 brain MRI images demonstrated that EFEDA achieves a Peak Signal-to-Noise Ratio (PSNR) improvement of 4.8 dB over the standard Canny filter. Furthermore, the Pratt's Figure of Merit scored 0.89, highlighting superior boundary preservation properties under 15% salt-and-pepper noise conditions. The visual results confirmed clearer delineation of micro-structures without introducing artificial artifacts, facilitating automated volumetric assessments of cerebral lesions.

Keywords: Image Segmentation, Fuzzy Logic, Edge Detection, Medical Imaging, MRI, Noise Reduction

Manuscript Timeline: Received July 05, 2010; Revised August 14, 2010; Accepted September 01, 2010; Published November 15, 2010.

Citation: Al-Farsi, T. A., & Tanaka, Y. (2010). An Enhanced Edge Detection Algorithm for Medical Image Segmentation Based on Fuzzy Logic. International Journal of Computer Science and Technology, 12(11), 76–82.

Subscribe to read the full article: https://www.internationalscholarsjournals.org/subscribe-to-read