Feature Extraction and Detection of Aorta using Histogram of Oriented Gradients and Support Vector Machines in Cardiac CTA

Authors

  • Ho Chul Kang

Abstract

In this paper, we propose an automatic aorta detection method in computed tomography angiography (CTA) using Histogram of Oriented Gradients (HOG) and Linear Support Vector Machine (LSVM). For our methods, we trained the LSVM classifier with HOG descriptors which are extracted from cardiac CTA. And we detect aorta region as follows. First, we denoise the images by applying anisotropic diffusion filtering. Second, the feature is extracted from the input image using HOG descriptor. Third, we detect the aorta by LSVM classifier. We tested our method in ten CT images and they were obtained from a different patient. For the evaluation of the computational performance of the proposed method, we measured the total processing time and intersection over union (IOU). The average of total processing time, from first step to third step, was 19.99±1.99s, and IOU was 0.85±0.05. We expect for our method to be used in cardiac diagnosis for cardiologist.

Downloads

Published

2020-03-23

Issue

Section

Articles