Analysis of Image Quality Assessment with Markov Random Field Oriented on Low Dose CT Images

In order to provide precision and objective image quality measures (IQM) for the low dose CT (Computed Tomography) images, various general IQMs need to be validated and analyzed. The IQM based on Markov Random Field (MRF) has not been check and validated by a comprehensive distorted database. First, this paper chooses a standard distorted image database of Laboratory for Image & Video Engineering (LIVE) to validate and analyze the performance of various IQMs. Then, the paper assesses various low dose phantom CT images by the IQMs. Experimental results show that the mutual information based on MRF is more obvious and precision than other measures to reflect the quality changes of the LIVE database and low dose CT phantom images. It can provide effective reference for quality assessment of low dose CT images.