The effect of a post-scan processing denoising system on image quality and morphometric analysis.

Area of Science:

• Neuroimaging
• Artificial Intelligence
• Medical Image Analysis

Background:

• Noise in Magnetic Resonance (MR) images significantly impacts image quality and the accuracy of brain morphometric analysis.
• Developing effective denoising techniques is crucial for reliable quantitative analysis of brain structure.

Purpose of the Study:

• To evaluate the efficacy of an AI-based denoising system, intelligent Quick Magnetic Resonance (iQMR), in enhancing MR image quality.
• To assess the impact of iQMR on the reliability and accuracy of brain morphometric analysis, specifically cortical thickness measurements.

Main Methods:

• Utilized 1.5T MP-RAGE MR images from the Alzheimer's Disease Neuroimaging Initiative 1 database.
• Compared MR images processed with and without the iQMR denoising system using both visual and objective assessments.
• Employed FreeSurfer for cortical thickness analysis in cross-sectional and longitudinal datasets (n=21 and n=15, respectively).

Main Results:

• The iQMR system significantly reduced image noise while preserving white-gray matter contrast.
• Intraclass correlation coefficients (ICCs) for cortical thickness were slightly improved with iQMR processing.
• Longitudinal analysis showed enhanced ICCs for symmetrized percent change and improved detectability of cortical thickness atrophy with iQMR.

Conclusions:

• The AI-based iQMR system effectively reduces noise and maintains contrast in MR images.
• iQMR processing demonstrably improves the reliability and detectability of brain morphometric analysis.
• This AI tool holds promise for enhancing quantitative neuroimaging studies.