Automated protocols for delineating human hippocampal subfields from 3 Tesla and 7 Tesla magnetic resonance imaging...

Area of Science:

• Neuroimaging
• Human Neuroscience
• Brain Anatomy

Background:

• Manual segmentation of human hippocampus subfields from MRI scans is time-consuming and requires specialized expertise.
• This limits research into the functional roles of distinct hippocampal subregions.
• Existing automated methods may lack the accuracy and validation needed for widespread adoption.

Purpose of the Study:

• To develop and validate automated protocols for segmenting the human hippocampus into six subregions.
• To provide protocols suitable for both 3T and 7T MRI data.
• To facilitate more efficient and accessible hippocampus subfield research.

Main Methods:

• Development of automated segmentation protocols trained on large cohorts of healthy young adults (n=140 at 3T, n=40 at 7T).
• Manual segmentation by expert raters was used as the gold standard for training and testing.
• Inter-rater reliability analyses were conducted to assess segmentation quality.

Main Results:

• The automated protocols achieved high-quality hippocampus subfield segmentation, comparable to expert manual segmenters.
• The protocols successfully segmented six distinct subregions: dentate gyrus/CA4, CA2/3, CA1, subiculum, pre/parasubiculum, and uncus.
• Open access to the protocols is provided for 3T and 7T MRI data.

Conclusions:

• The developed automated protocols significantly reduce the time and expertise required for hippocampus subfield segmentation.
• These validated tools are expected to accelerate research into hippocampus function and its role in various cognitive processes.
• The open-access nature of the protocols promotes wider adoption and advancement in the field.