Related Experiment Video
Updated: May 29, 2026
Automated Segmentation of Cortical Grey Matter from T1-Weighted MRI Images
Published on: January 7, 2019
Validation of automated white matter hyperintensity segmentation
Sean D Smart1, Michael J Firbank, John T O'Brien
1Institute for Ageing and Health, Newcastle University, Campus for Ageing and Vitality, Newcastle upon Tyne NE4 5PL, UK.
Abstract:
Introduction. White matter hyperintensities (WMHs) are a common finding on MRI scans of older people and are associated with vascular disease. We compared 3 methods for automatically segmenting WMHs from MRI scans. Method. An operator manually segmented WMHs on MRI images from a 3T scanner. The scans were also segmented in a fully automated fashion by three different programmes. The voxel overlap between manual and automated segmentation was compared. Results. Between observer overlap ratio was 63%. Using our previously described in-house software, we had overlap of 62.2%. We investigated the use of a modified version of SPM segmentation; however, this was not successful, with only 14% overlap. Discussion. Using our previously reported software, we demonstrated good segmentation of WMHs in a fully automated fashion.
Insights
Automated segmentation of white matter hyperintensities (WMHs) using in-house software achieved 62.2% overlap, comparable to manual segmentation. This demonstrates a reliable method for analyzing these common MRI findings in older adults.
Area of Science:
- Neuroimaging
- Medical image analysis
- Radiology
Background:
- White matter hyperintensities (WMHs) are prevalent in older adults' MRI scans.
- WMHs are frequently linked to cerebrovascular disease.
- Accurate segmentation of WMHs is crucial for research and clinical applications.
Purpose of the Study:
- To compare the accuracy of three automated methods for segmenting white matter hyperintensities (WMHs) against manual segmentation.
- To evaluate the performance of a previously developed in-house software for WMH segmentation.
- To assess the feasibility of using modified SPM segmentation for WMH analysis.
Main Methods:
- Manual segmentation of WMHs by an operator on 3T MRI scans.
- Fully automated segmentation using three distinct software programs.
- Quantitative comparison of automated segmentations with manual segmentation using voxel overlap.
Main Results:
- The in-house software achieved a 62.2% voxel overlap, closely matching the 63% between-observer manual segmentation overlap.
- Modified SPM segmentation showed poor performance with only 14% overlap.
- The in-house software demonstrated good agreement with manual segmentation.
Conclusions:
- The previously reported in-house software provides a reliable and automated method for segmenting white matter hyperintensities.
- Automated WMH segmentation using this software can aid in the study of cerebrovascular disease in aging populations.
- Further development of automated segmentation tools is essential for efficient neuroimaging analysis.
