Computational Method to Extract the Keyframe from Angiogram video

Abstract

Extraction of keyframes from the coronary angiogram video is an essential task in the medical community to analyze the image and proceed for further treatment. There are different medical modalities such as MRI, CT, X-ray, etc. to capture the images.  This research focuses on coronary artery disease which is identified by the X-ray coronary angiogram test. The process of acquiring images in X-ray angiogram data is generally capturing the video at 15-30 frames per second. During the analysis of the coronary angiogram video the cardiologist often manually pauses the video at a regular interval frame to examine the artery. Such a frame which represents the vessels or artery with complete visual content is a keyframe. In this paper a method to extract the keyframe automatically from the angiogram video is proposed.  This proposed method has three phases, In the first phase video is converted into frames. In the second phase frames are pre-processed to improve the quality of the frames. In the third phase, by comparing connected components properties of each frame, keyframes are extracted. This keyframe is used for further processes like segmentation and detection of the stenosis in the artery and the results of the proposed method is compared with the existing method.