Title: Modified fuzzy clustering-based segmentation through histogram combined with K-NN classification

Authors: Balan Thamaraichelvi

Addresses: Department of Electrical Engineering, FEAT, Annamalai University, Tamil Nadu, India

Abstract: Medical image analysis plays a vital role in diagnosing the disease accurately in the medical field. Image segmentation is a challenging problem in the field of medical image analysis. In this paper, a modified Gaussian kernelised additive bias field clustering-based segmentation technique with unsupervised K-NN classification technique has been considered to analyse the magnetic resonance (MR) brain images for tissue segmentation and tumour detection. The accuracy of the proposed segmentation and classification techniques is found to be around 95%. The accuracy and the statistical measures like selectivity and sensitivity are calculated using the fractions: true negative (TN), true positive (TP), false negative (FN) and false positive (FP).

Keywords: image segmentation; histogram-based centre initialisation; fuzzy C-means; FCM; Gaussian radial basis kernel function; discrete wavelet transform; DWT; principal component analysis; PCA; K-NN classification.

DOI: 10.1504/IJMEI.2021.117730

International Journal of Medical Engineering and Informatics, 2021 Vol.13 No.5, pp.410 - 418

Received: 18 Jul 2019
Accepted: 14 Dec 2019
Published online: 23 Sep 2021 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article