Ultra-low-field brain MRI morphometry: Test-retest reliability and correspondence to high-field MRI

Area of Science:

• Neuroimaging
• Biomedical Engineering
• Radiology

Background:

• Magnetic resonance imaging (MRI) is crucial for brain monitoring but high-field systems are costly and resource-intensive.
• Ultra-low-field (ULF) MRI (<0.1 T) offers a more accessible and affordable alternative.
• The reliability of ULF MRI for quantitative brain morphometry and its correspondence with high-field MRI remain uncertain.

Purpose of the Study:

• To evaluate the reliability of anatomical measurements from ULF MRI.
• To assess the correspondence between ULF MRI morphometry and high-field MRI data.
• To explore the potential of ULF MRI for studying neuroanatomy across the lifespan and in disease.

Main Methods:

• Scanned 23 healthy adults (20-69 years) using two 64 mT ULF MRI systems and one 3 T high-field MRI system.
• Acquired T1w and T2w scans at various resolutions on ULF systems.
• Segmented brain images into global tissue types and local structures, analyzing morphometric reliability and correspondence using volume correlations and Dice overlap.

Main Results:

• Demonstrated high reliability for 64 mT morphometry and strong correspondence with 3 T measurements across different contrasts and resolutions.
• Achieved optimal performance by combining multiple low-resolution T2w scans into a higher-resolution image via multi-resolution registration.
• Observed that larger brain structures exhibit higher reliability and better correspondence with high-field MRI.

Conclusions:

• ULF MRI provides reliable quantitative brain morphometry comparable to high-field MRI.
• This technology holds significant potential for accessible neuroimaging across the lifespan and for monitoring neurological conditions.
• Publicly available raw data facilitates validation of ULF neuroimaging analysis techniques.