HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Services Email this article to a friend Similar articles in this journal Alert me to new issues of the journal Download to citation manager Google Scholar Articles by Bagher-Ebadian, H. Articles by Setayeshi, S. PubMed Articles by Bagher-Ebadian, H. Articles by Setayeshi, S.

A hierarchical fuzzy clustering model for myocardial segmentation in cardiac SPECT images

¹ Department of Nuclear Engineering & Physics, Amir-Kabir University of Technology, Tehran, Iran; ² Control and Intelligent Processing Center of Excellence, University of Tehran, Tehran, Iran

496

Objectives: Image information extraction is one of the most important aspects in both therapeutic and diagnosis fields of nuclear medicine. In this study, a Hierarchical Fuzzy Clustering (HFC) model is designed to segment the myocardial from its background in Cardiac SPECT images. The HFC model was constructed from a set of interconnected FC units to segment the myocardial structure in sequential stages.The model was evaluated and tested using an analytical cardiac SPECT simulation method and SPECT images from real patients.

Methods: In this model, each FC unit was fed by the histogram (count) of its previous unit and new histogram of the segmented information was passed to the next unit. Therefore, the model segmented the myocardial data in a few stages by shifting the threshold value toward the hot spot area (high count area). Each FC unit employs the Picard iteration algorithm to converge and the optimal number of units was found by segmentation of simulated(20)and real cases(40).

Results: Results imply that three FC units can adequately segment the myocardial images in both simulation and real cases. To evaluate the model, it was applied to all simulated and SPECT images of real patients (ROIs were drawn by two experts). The average percentage of the absolute error between the measure of segmented volume for both simulated and real cases (%15 and %19 for simulated and real cases respectively) was calculated.

Conclusions: A HFC Model for Myocardial Segmentation was introduced and evaluated. Results imply that the model has a good potential for volume segmentation of the myocardial structure in Cardiac SPECT Images.

Research Support: This work was supported by Amir-Kabir University of Technology, Tehran, Iran.

This Article Services Email this article to a friend Similar articles in this journal Alert me to new issues of the journal Download to citation manager Google Scholar Articles by Bagher-Ebadian, H. Articles by Setayeshi, S. PubMed Articles by Bagher-Ebadian, H. Articles by Setayeshi, S.