American Journal of Neuroradiology, Vol 19, Issue 9 1715-1720, Copyright © 1998 by American Society of Neuroradiology
ARTICLES |
The effect of section thickness on MR lesion detection and quantification in multiple sclerosis
PD Molyneux, N Tubridy, GJ Parker, GJ Barker, DG MacManus, PS Tofts, IF Moseley and DH Miller
NMR Research Unit, Institute of Neurology, London, United Kingdom.
BACKGROUND AND PURPOSE: The purpose of our study was to investigate the effect of section thickness on MR detection of brain lesion volume and measurement precision in patients with multiple sclerosis (MS). METHODS: Eight subjects with known MS were studied on a 1.5-T MR system. We used a 3D fast fluid-attenuated inversion-recovery sequence to obtain contiguous axial brain images at section thicknesses of 5 mm, 3 mm, and 1 mm. Two sets of images were acquired at each section thickness during two sessions, between which the patient was removed from the scanner. Lesion volumes were measured at each section thickness using a semiautomated local thresholding technique. RESULTS: We found that progressive reduction in section thickness led to detection of smaller lesions, resulting in a significant (8%) increase in lesion volume on MR images as section thickness was reduced from 5 mm to 3 mm. However, despite a further increase in lesion detection at a section thickness of 1 mm, this did not result in an increase in total lesion volume. This finding indicates that the relationship between section thickness and lesion volume on MR images is not linear. Scan-rescan reproducibility was improved by reducing section thickness, at the cost of increased analysis time. CONCLUSION: This study shows that acquisition of very thin sections increases the sensitivity and precision of MS lesion measurement. Serial studies assessing lesion changes over time are needed to define the impact of this increase on sample size requirements for MS treatment trials.