Neuroscience 2001 Abstract
| Presentation Number: | 966.3 |
|---|---|
| Abstract Title: | Cerebral volumetry toolbox with iterative 3D region growing to determine segmentation threshold levels for semiautomatic MRI-volumetry. |
| Authors: |
Hauser, T. K.*1
; Luft, A. R.1
; Skalej, M.2
; Schulz, J. B.1
1Dept Neurol, Univ Tuebingen, Tuebingen, Germany 2Dept Neuroradiol, Univ Tuebingen, Tuebingen, Germany |
| Primary Theme and Topics |
Neurological and Psychiatric Conditions - Neurodegenerative Disorders -- Parkinsons Disease: Other |
| Session: |
966. Neurodegenerative disorders: Parkinson's disease--clinical Poster |
| Presentation Time: | Thursday, November 15, 2001 10:00 AM-11:00 AM |
| Location: | Exhibit Hall VV-61 |
| Keywords: | NEURODEGENERATION, CEREBRAL LOBES, VOLUMETRY, SEGMENTATION |
A novel procedure to measure the volume of gray and white matter in each lobe of the cerebral hemispheres was developed. Semi-automatic segmentation and volumetry is performed using high-resolution MR image sets. While previous techniques relied on manual tracing or segmentation in Talairach space, this procedure uses anatomical landmarks combined with region growing to improve reliability and to reduce analysis time.
The procedure was implemented as a MATLAB toolbox. First, the skull was removed from the MRI volume before brain stem and cerebellum were separated as previously published. On the resulting volume comprising hemispheres and diencephalon, a novel iterated 3D region growing algorithm was applied to determine GM/WM intensity thresholds. This algorithm uses the derivative of the function between threshold and voxel count to determine optimal threshold values. It thereby minimizes subjective rater influence. The hemispheres were then warped to Talairach space. Lobar boundaries derived from the Talairach-atlas were manually adapted to the individual gyral pattern. Lobar volumes of gray and white matter were then calculated automatically. Lobar volumes in nine subjects were determined by two independent raters. The intra-class correlation was 0.9726 (p<0,0001) for GM and 0,9785 (p<0.0001) for WM. Mean duration of the entire procedure was 120 ± 20 minutes.
The procedure allows for fast assessment of lobar brain volumes with high inter-rater reliability. It can be applied to study brain atrophy in various neurodegenerative disease states. Easy and fast use may allow its adoption in clinical routine.
The procedure was implemented as a MATLAB toolbox. First, the skull was removed from the MRI volume before brain stem and cerebellum were separated as previously published. On the resulting volume comprising hemispheres and diencephalon, a novel iterated 3D region growing algorithm was applied to determine GM/WM intensity thresholds. This algorithm uses the derivative of the function between threshold and voxel count to determine optimal threshold values. It thereby minimizes subjective rater influence. The hemispheres were then warped to Talairach space. Lobar boundaries derived from the Talairach-atlas were manually adapted to the individual gyral pattern. Lobar volumes of gray and white matter were then calculated automatically. Lobar volumes in nine subjects were determined by two independent raters. The intra-class correlation was 0.9726 (p<0,0001) for GM and 0,9785 (p<0.0001) for WM. Mean duration of the entire procedure was 120 ± 20 minutes.
The procedure allows for fast assessment of lobar brain volumes with high inter-rater reliability. It can be applied to study brain atrophy in various neurodegenerative disease states. Easy and fast use may allow its adoption in clinical routine.
Sample Citation:
[Authors]. [Abstract Title]. Program No. XXX.XX. 2001 Neuroscience Meeting Planner. San Diego, CA: Society for Neuroscience, 2001. Online.
Copyright © 2001-2026 Society for Neuroscience; all rights reserved. Permission to republish any abstract or part of any abstract in any form must be obtained in writing by SfN office prior to publication.
