Accuracy of automated amygdala MRI segmentation approaches in Huntington's disease in the IMAGE-HD cohort
Related Experiment Video
Updated: Dec 29, 2025
Whole-brain Segmentation and Change-point Analysis of Anatomical Brain MRI—Application in Premanifest Huntington's Disease
Published on: June 9, 2018
Bonnie Alexander1,2, Nellie Georgiou-Karistianis1, Richard Beare2,3
1Turner Institute for Brain and Mental Health, School of Psychological Sciences, Monash University, Melbourne, Victoria, Australia.
Abstract:
Smaller manually-segmented amygdala volumes have been associated with poorer motor and cognitive function in Huntington's disease (HD). Manual segmentation is the gold standard in terms of accuracy; however, automated methods may be necessary in large samples. Automated segmentation accuracy has not been determined for the amygdala in HD. We aimed to determine which of three automated approaches would most accurately segment amygdalae in HD: FreeSurfer, FIRST, and ANTS nonlinear registration followed by FIRST segmentation. T1-weighted images for the IMAGE-HD cohort including 35 presymptomatic HD (pre-HD), 36 symptomatic HD (symp-HD), and 34 healthy controls were segmented using FreeSurfer and FIRST. For the third approach, images were nonlinearly registered to an MNI template using ANTS, then segmented using FIRST. All automated methods overestimated amygdala volumes compared with manual segmentation. Dice overlap scores, indicating segmentation accuracy, were not significantly different between automated approaches. Manually segmented volumes were most statistically differentiable between groups, followed by those segmented by FreeSurfer, then ANTS/FIRST. FIRST-segmented volumes did not differ between groups. All automated methods produced a bias where volume overestimation was more severe for smaller amygdalae. This bias was subtle for FreeSurfer, but marked for FIRST, and moderate for ANTS/FIRST. Further, FreeSurfer introduced a hemispheric bias not evident with manual segmentation, producing larger right amygdalae by 8%. To assist choice of segmentation approach, we provide sample size estimation graphs based on sample size and other factors. If automated segmentation is employed in samples of the current size, FreeSurfer may effectively distinguish amygdala volume between controls and HD.
Insights
Automated brain imaging segmentation in Huntington's disease (HD) shows FreeSurfer may be best for amygdala volume analysis, despite overestimation bias. Manual segmentation remains most accurate for group differences.
Area of Science:
- Neuroimaging
- Neurodegeneration
- Computational Anatomy
Background:
- Smaller amygdala volumes correlate with poorer motor and cognitive function in Huntington's disease (HD).
- Manual segmentation is accurate but time-consuming for large cohorts.
- Automated segmentation accuracy for the amygdala in HD is not well-established.
Purpose of the Study:
- To evaluate the accuracy of three automated brain imaging segmentation methods for the amygdala in Huntington's disease.
- To compare FreeSurfer, FIRST, and ANTS nonlinear registration followed by FIRST segmentation for amygdala volume analysis in HD.
- To identify the most suitable automated method for large-scale HD neuroimaging studies.
Main Methods:
- T1-weighted MRI data from 35 presymptomatic HD, 36 symptomatic HD, and 34 healthy controls (IMAGE-HD cohort) were analyzed.
- Amygdalae were segmented using FreeSurfer, FIRST, and a combination of ANTS nonlinear registration with FIRST.
- Automated segmentations were compared against manual segmentations using Dice overlap scores and statistical group differentiability.
Main Results:
- All automated methods overestimated amygdala volumes compared to manual segmentation.
- No significant differences in Dice overlap scores were found between the automated methods.
- Manual segmentation showed the greatest statistical differentiation between groups, followed by FreeSurfer, then ANTS/FIRST; FIRST alone did not differentiate groups.
- Automated methods exhibited a bias of increased overestimation for smaller amygdalae, most pronounced in FIRST and moderate in ANTS/FIRST, but subtle in FreeSurfer.
- FreeSurfer introduced an 8% hemispheric bias, overestimating right amygdala volume compared to manual segmentation.
Conclusions:
- FreeSurfer demonstrated the most favorable performance among the tested automated methods for amygdala segmentation in HD, despite inherent biases.
- Automated segmentation, particularly FreeSurfer, may be viable for distinguishing amygdala volume differences in large HD cohorts.
- Careful consideration of segmentation method biases and accuracy is crucial for reliable neuroimaging research in neurodegenerative diseases like HD.
