A radiographic image quality assessment algorithm based on network topology analysis

摘要

To measure enhanced RT image quality, a novel reduced-reference image quality assessment (RR-IQA) method for RT images, based on measuring the topology property of the RT image network, is introduced in this paper. We suppose that the original RT image is perfect, and the enhanced images are similar to the perfect image. In this paper, we construct a network model of RT images and analyse the topology property of the RT image networks. Then, we use the topology property to represent the similarity between the reference RT image and the enhanced RT image. The robustness of the method has been experimentally validated in the case of 150 enhanced RT images. Experimental results indicate that the objective scores obtained by the proposed method correlate well with the subjective assessment scores, and the proposed method is better than recent state-of-the-art image quality assessment methods.