Medical Image Quality Assessment Using CSO Based Deep Neural Network

A deep learning approach to image quality assessment

“…Hua et al[180] propose a structural similarity based full-reference MIQA specifically for assessing quality of MR images. Jayageetha et al[181] propose a CNN based no-reference MIQA and evaluate it with four publically available medical image databases. Instead of designing specific image modalities, Zhang et al[182] explore whether off-the-shelf no-reference IQA methods designed for natural images can predict the quality of MR images, and conclude that BLIINDS-II[68] and BRISQUE[69] are full of potential to be re-purposed.…”