Precision brain morphometry using cluster scanning

Area of Science:

• Neuroimaging
• Radiology
• Biomarkers

Background:

• Measurement error in structural MRI limits detecting group differences and longitudinal changes.
• Accelerated MRI sequences (~1 minute) offer morphometric errors comparable to traditional scans.

Purpose of the Study:

• To evaluate the efficacy of cluster scanning for improving brain morphometric precision.
• To assess the impact of pooling estimates from rapid MRI acquisitions on measurement error.

Main Methods:

• A test-retest study involving 40 individuals across different age and cognitive groups.
• Acquisition of single and clustered compressed sensing (CS) T1-weighted MRI scans.
• Comparison of morphometric errors from CS scans with a traditional Alzheimer's Disease Neuroimaging Initiative (ADNI) protocol.

Main Results:

• Single CS scans showed ~12% larger morphometric errors than ADNI scans.
• Pooling estimates from four clustered CS acquisitions reduced errors by 34% compared to ADNI.
• Pooling estimates from eight CS scans further reduced errors by 51% compared to ADNI.

Conclusions:

• Cluster scanning with accelerated MRI significantly enhances measurement precision in brain morphometry.
• This approach offers a method to improve statistical power for detecting neurodegenerative changes.
• Cluster scanning holds potential for developing more sensitive biomarkers of disease progression.