Automatic analysis of skull thickness, scalp‐to‐cortex distance and association with age and sex in cognitively normal elderly

Abstract

AbstractBackgroundTranscranial brain stimulation has been a potential treatment for many brain diseases. The skull compartment exerts a strong influence on the delivery of stimulation, implying the need to accurately examine its geometry and alterations in the population. In addition, personalized neurostimulation requires better understanding of skull geometry and scalp‐to‐cortex distance for optimization.MethodHere, we developed an open source efficient pipeline BrainCalculator based on BrainSuite Skullfinder for automatically computing the skull thickness map to locate the position of minimum thickness, SCD, and brain volume based on structural T1‐weighted MRI 3T data. We examined the influence of age and sex cross‐sectionally in 407 cognitively normal older adults (71.93±7.98 years, 60.2% female) from the ADNI on the minimum skull thickness, SCD and brain volume of the whole‐head MR data. We demonstrated the compatibility of BrainCalculator with available preprocessing toolboxes, including FSL Brain Extraction Tool (BET) 2 and Statistical Parametric Mapping (SPM)12‐based unified segmentation, and compared their performance.ResultOur pipeline was able to compute automatically the skull thinkness, SCD and brain volume using T1 MR data in 8 minutes. We found that the sphenoid bone and temporal bone were thinnest among the skull regions. There was no increase in minimum skull thickness at different regions with increasing age except in the female sphenoid bone. No sex difference in minimum skull thickness or SCD in different age groups was observed. Positive correlations between age and SCD in both males and females were observed, and the slope was higher in the temporal cortex SCD in females (0.307%/y) than in males (0.216%/y) but comparable in other regions. A negative correlation between age and whole brain volume computed based on brain surface in both males and females was observed, and the slope was slightly higher in females (‐1.031%/y) than in males (‐0.998%/y).ConclusionWe developed a pipeline for MR‐based skull thickness and SCD brain volume automatic analysis and demonstrated the sex‐dependent association between minimum regional skull thickness, SCD and brain volume with age. This pipeline is potentially useful for personalized neurostimulation.