Abdominal Aortic ANEURYSM Identification Using HLSFMM Segmentation and SVM Classifier

Abstract

The localized inflammation of the abdominal aorta region causes Abdominal Aortic Aneurysm (AAA). The width of the lumen enlarges its size 3 cm or more than half of its diameter, which is larger than the typical diameter. There is no symptom until it becomes ruptured, which may often results in death. In this paper, a hybrid level set technique is presented to detect and segment the image taken from MRI of abdominal aortic aneurysm region. In traditional level set technique re-initialization problems are high. This problem is completely eradicated in the Hybrid Level Set Fast Marching method (HLSFMM). Median filter diminishes the noise in the image efficiently when compared to standard SVM classifier which uses Gaussian RBF kernel operator as a diameter measure by incorporating spatial data. Finally HLSFMM is utilized to extract source boundary in pre segmentation stage. The precision and the orderliness of the proposed method are extracted for different noisy MRI AAA images. Compared this result with other methods, the proposed system is much proficient for images with noises and accurate segmentations results are attained.