How accurate are measurements on MRI? A study on multiple sclerosis using reliable 3D stereological methods

Area of Science:

• Neurology
• Medical Imaging
• Quantitative Anatomy

Background:

• Multiple sclerosis (MS) is characterized by demyelination and axonal damage, leading to the formation of plaques in the brain.
• Accurate quantification of MS plaque load is crucial for understanding disease progression and evaluating treatment efficacy.
• Current methods for estimating plaque volume and number may be subject to bias.

Purpose of the Study:

• To describe unbiased stereological principles for quantifying plaque volume and number in MS using MRI.
• To discuss practical challenges in implementing these stereological methods.
• To assess the reliability and efficiency of stereological methods for MS plaque load estimation.

Main Methods:

• Estimation of plaque volume using stereological methods on 3 mm brain MRI scans.
• Application of a bias correction for convex particles (plaques) by excluding the slice of maximal area.
• Utilization of the disector, an unbiased three-dimensional counting rule, for plaque number estimation.
• Calculation of the coefficient of error for the estimates.

Main Results:

• Volume estimates from thick sections (3 mm) were found to be biased, with bias ranging from 15% to 200% depending on object size.
• Bias in plaque volume estimation varied between 15% and 90% when applying corrections.
• Using 1 mm slices significantly reduced bias to near zero, making them preferable for precise measurements.
• Stereological methods provided efficient and reliable estimates of MS plaque load.

Conclusions:

• Stereological principles, when applied to MRI data, offer an unbiased approach to quantifying MS plaque volume and number.
• The use of 1 mm slices is recommended for accurate plaque volume estimation.
• Stereological methods provide a robust framework for assessing MS disease burden through MRI.