Cortical thickness analysis examined through power analysis and a population simulation.

Area of Science:

• Neuroimaging
• Computational Neuroscience
• Medical Imaging Analysis

Background:

• Accurate measurement of cerebral cortex thickness is vital for understanding brain development and disease.
• Previous development of an automated surface-extraction (ASP) algorithm for whole-brain cortical thickness measurement.
• Need to assess the precision, optimal parameters, and sensitivity of this automated method.

Purpose of the Study:

• To evaluate the precision and reliability of the automated surface-extraction (ASP) algorithm for cortical thickness measurement.
• To determine the optimal performance parameters for the ASP algorithm.
• To assess the sensitivity of the method to detect subtle, focal changes in cortical thickness.

Main Methods:

• Utilized 3-D MRI data and a fully automated surface-extraction (ASP) algorithm.
• Assessed precision through simulated population studies and single-subject reproducibility metrics.
• Compared six different cortical thickness metrics, identifying the most precise method as the distance between white and gray matter surfaces.

Main Results:

• The automated cortical thickness measurement method demonstrated high reliability and precision.
• Achieved a sensitivity of 0.93 (probability of a true-positive) for detecting changes.
• A 0.6-mm (15%) change in thickness was detectable in groups of 25 subjects after applying a 3-D Gaussian kernel (FWHM = 30 mm).
• Optimal precision was achieved with smoothing across the 2-D surface manifold using a 30 mm kernel size.

Conclusions:

• The automated surface-extraction (ASP) algorithm provides a precise and reliable method for measuring whole-brain cortical thickness from 3-D MRI data.
• The method is sensitive enough to detect subtle, focal changes in cortical thickness, with optimal parameters identified for enhanced accuracy.
• This technique holds significant potential for advancing neurological research and clinical diagnostics by enabling quantitative analysis of cortical morphology.