Volume 20, Issue 2 (June 2024)                   IJEEE 2024, 20(2): 32-41 | Back to browse issues page


XML Print


Abstract:   (727 Views)
Cardiovascular diseases (CVD) are today a major cause of death globally that is diagnosed by measurement and quantification of left ventricle (LV) wall motion (WM) abnormality of the heart. The aim of this study was to assess the utility of left ventricular (LV) entropy, a novel measure of disease derived from two-dimensional (2D) echocardiography images that assesses the probability distribution of pixel intensities in the LV. The purpose of this research is to develop the method of LV entropy to predict heart diseases. In this algorithm, a frame is usually chosen as the reference frame to extract the region of interest (ROI) around LV and then it is mapped to all images in a cardiac cycle. Then Shannon Entropy transform was applied to calculate the distribution of pixel intensities across the LV so we obtained entropy curves and compared them. The main idea is to find a motion estimation accuracy. The results obtained by our method are quantitatively evaluated to those obtained by an experienced echocardiographer visually on 22 normal cases and 19 myocardial infarction (MI) cases in apical four-chamber (A4C) view. The entropy of diastole in MI cases was 0.50 (0.29-0.58) while in normal cases was 0.75 (0.64-1.13). The entropy of systole in MI cases was 0.64 (0.26-1.04) while in normal cases was 0.81 (0.63-1.26). The percent change of entropy for diastole and systole between normal and MI cases are 33.3% and 20.2%. The results indicate that the LV entropy curves of MI cases have less changes than normal cases.
Full-Text [PDF 617 kb]   (235 Downloads)    
Type of Study: Research Paper | Subject: Biomedical Signal & Image Processing
Received: 2023/08/13 | Revised: 2024/08/31 | Accepted: 2024/05/27

Rights and permissions
Creative Commons License This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.