Paper Info
Title
Lumen boundary detection in IVUS medical imaging using structured element.
Abstract
The lumen boundary in the human coronary artery is the contour edge of a blood vessel. The intravascular ultrasound (IVUS) is the medical imaging modality used to view the lumen boundary by the clinicians to detect the coronary artery disease called atherosclerosis. The main problem is to differentiate between the lumen area and the lumen boundary which cannot see clearly. The diameter of the lumen becomes narrowed because of the plaques, lipids and calcium deposits on the artery wall. In this paper, we present the automated segmentation method for detecting the lumen boundary using Otsu threshold, morphological operation and empirical threshold in the IVUS images. We used six types of structured elements to select the best result for automated segmentation of lumen boundary. Forty samples of IVUS images inclusive of the ground truth obtained from the Universitat de Barcelona, Barcelona used in this study. The proposed method segmentation performance measured are Jaccard-Index, Dice Similarity-Index, Hausdorff-Distance, Area Overlapped Error and Percentage Area Difference. The Bland-Altman Plot is used to show the variation between the proposed automatic segmentation area and ground truth area. The structured element of the octagon gave a good result in Hausdorff Distance, and the line gave a better result in Jaccard Index, Percentage Area Distance, Area Overlapped Error, Dice Index and Area Error. The result obtained shows that the segmentation performance of the proposed method is on par with other existing segmentation methods.
Year
DOI
Venue
2017
10.1145/3022227.3022296
IMCOM
Field
DocType
Citations 
Computer vision,Intravascular ultrasound,Medical imaging,Segmentation,Computer science,Lumen (unit),Boundary detection,Ground truth,Artificial intelligence,Hausdorff distance,Jaccard index
Conference
0
PageRank 
References 
Authors
0.34
3
4
Name
Order
Citations
PageRank
Hannah Sofian100.34
Joel C. Than2262.52
Norliza Mohd. Noor3379.25
Suraya Mohamad400.34