Abstract:

The problem of segmentation and annotation of medical images is the difficulty and hotspot of current medical image processing research.At present, there is no unified definition and consistent annotation method for manual annotation of coronary CT medical images.Therefore, the image annotation method of coronary CT angiography is discussed according to the position of organs in medical images.This paper discusses the annotation format, principles and strategies of coronary CT angiography.According to the segmented image annotation principle, the whole coronary CT angiography image is subdivided into twelve segments.Combined with specific examples, the specific annotation methods of segmented images of right coronary artery branch and right posterior descending branch are expounded, that is, selective annotation, how to annotate and annotation form, etc.This segmented annotation method of CT angiography for organ positions in medical images can build an efficient and unified training data set for automatic annotation of medical images, and improve the efficiency of automatic image annotation based on supervised learning or semi-supervised learning.

Key words: Coronary artery, CT angiography, CT image annotation, Medical image processing