A Unified Automated Segmentation Technique for the Left Ventricle Segmentation in Cardiac MRI

• Published in: IOP conference series. Materials Science and Engineering 2020-12, Vol.993 (1), p.12068
• Main Authors: Sivakumar, K., Lavanya, S., Himanandini, Karna, Keerthana, V.
• Format: Article
• Language: English
• Subjects: APD; Bias field Correction; DICE; Energy pre-fitting; Intensity inhomogeneity; Level set; OTSU threshold
• ISSN: 1757-8981; 1757-899X
• DOI: 10.1088/1757-899X/993/1/012068

Though the technology updated for every minute of time to aid all kinds of problems facing in every day practical life, the technique requires amelioration to solve or support specific kinds of problems like early identification of cardiovascular disease. For the past decades, cardiovascular diseases (CVD) are the major reason for death. Cardiac MRI is useful for acquiring the anatomical data of the alive heart for the clinical diagnosis of cardiovascular diseases. F rom the LV Segmentation on cardiac MRI, the important parameters estimated to diagnose are ejection fraction, left ventricle myocardium mass, stroke volume, etc. Hence segmenting the LV automatically plays a significant role in helping the physician to test cardiac functions quickly since manual segmentation is a time-consuming work. This Automatic Segmentation also eliminates manual errors during evaluation. Therefore, to discuss the problem we propose enhanced techniques that is a unified method to segment the left ventricle automatically from the cardiac MRI image. The algorithm demonstrated in this work is to combine existing segmentations in an efficient way to give even more efficient and accurate results. two performance evaluation techniques namely APD and DICE to quantify the result. Outcomes obtained are then compared with recent segmentation methods to show the efficiency of the unified segmentation technique.