Performance Assessment of K-Means and KNN Algorithm for the Detection of Borders and Extraction of Plaque Area from IVUS Images
C. Mahadevi1, S. Sivakumar2
1C. Mahadevi*, Assistant Professor, Department of Computer Science, N.M.S.S.Vellaichamy Nadar College, Nagamalai, Madurai.
2Dr. S. Sivakumar, Associate Professor and Head, Department of Computer Science, Cardamom Planters’ Association College, Bodinayakanur, Theni District, TN, India. 

Manuscript received on January 05, 2020. | Revised Manuscript received on January 25, 2020. | Manuscript published on January 30, 2020. | PP: 4990-4998 | Volume-8 Issue-5, January 2020. | Retrieval Number: E6853018520/2020©BEIESP | DOI: 10.35940/ijrte.E6853.018520

Open Access | Ethics and Policies | Cite  | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Abstract: The coronary artery vascular disease of atherosclerosis, which made the blood vessel artery wall harden and narrow. The vascular wall disease quantitatively analyzed and diagnosed by an intravascular ultrasound (IVUS) image. The quantitative investigations of coronary atherosclerosis by means of IVUS and manual recognition of wall and plaque borders are restricted by the need for observers with considerable understanding and the tedious environment of manual border detection. To improve and provide more detailed vessel and plaque information for better diagnosis and assessment go for an automated segmentation. An automated construction for the purpose of perceiving lumen and media-adventitia borders in IVUS images was effectively formulated. An effectual unsupervised K-Means clustering scheme refine the borders with morphological operations, later an IVUS data samples area classified using supervised KNN(K-Nearest Neighbor) classifier to extract the plaque feature. The performance of contour metric measurements in terms of Jaccard Index (JI), percentage area difference (PAD), area error (AE), dice index (DI), false positive ratio (RFP), in addition to false negative ratio Morphology (RFN) are computed for evaluation and variation. 

Keywords: Morphology Operation, Region Props, K-Means Clustering, KNN Classifier.
Scope of the Article: Network Operations & Management.