TIQA-MRI: Toolbox for Perceptual Image Quality Assessment of Magnetic Resonance Images

Magnetic Resonance Imaging (MRI) plays a pivotal role in medical diagnostics and research as a non-invasive imaging tool. The accuracy and reliability of clinical evaluations depend heavily on the quality of MRI images, making high-quality imaging indispensable. Unfortunately, many MRI scans are compromised by noise, artifacts, or distortions, highlighting the need for no-reference Image Quality Assessment (IQA) methods to evaluate MR image quality. The proposed Toolbox for Perceptual Image Quality Assessment of Magnetic Resonance Images (TIQA-MRI) addresses this issue by offering a dedicated platform for assessing MR image quality. TIQA-MRI supports the development, testing, and evaluation of MR-IQA techniques through a collection of objective IQA methods and a curated dataset of MR images with subjective scores, facilitating comprehensive quality assessment. The toolbox allows users to evaluate the quality of their images and compare their MR-IQA techniques with existing methods. Implemented in MATLAB and available on GitHub, TIQA-MRI offers a standardized and accessible framework to advance research in MR image quality assessment.