Related Experiment Video
Updated: Jul 30, 2025
08:57
Quantification of Optic Nerve Cross Sectional Area on MRI: A Novel Protocol using Fiji Software
Published on: September 4, 2021
4.1K
Automatic segmentation and quantification of the optic nerve on MRI using a 3D U-Net
Sabien van Elst1,2, Christiaan M de Bloeme1,2, Samantha Noteboom3
1Amsterdam UMC location Vrije Universiteit Amsterdam, Department of Radiology and Nuclear Medicine, Amsterdam, The Netherlands.
Journal of Medical Imaging (Bellingham, Wash.)
|May 17, 2023
Summary
This study developed an automated framework using MRI to segment the optic nerve (ON) and quantify its structural changes. The method accurately measures ON diameter and area, aiding in understanding disease mechanisms.
Area of Science:
- Neuroimaging
- Medical image analysis
- Ophthalmology
Background:
- Optic nerve (ON) pathologies can cause structural changes, impacting disease understanding.
- Accurate quantification of ON structural changes is crucial for disease mechanism research.
Purpose of the Study:
- To develop an automated framework for segmenting the optic nerve (ON) from cerebrospinal fluid (CSF) in MRI.
- To quantify ON diameter and cross-sectional area along its entire length using automated segmentation.
Main Methods:
- Utilized a 3D U-Net for automated ON segmentation on 40 T2-weighted MRI scans.
- Employed centerline extraction and 3D surface models for ON quantification.
- Assessed performance using Dice similarity coefficient, Hausdorff distance, and intraclass correlation coefficient (ICC).
Main Results:
- The segmentation network achieved a mean Dice score of 0.84 and ICC of 0.95 on the test set.
- Quantification showed mean ICC of 0.76 for diameter and 0.71 for cross-sectional area.
- The method accurately differentiates ON from CSF and estimates diameter along the ON centerline.
Conclusions:
- An automated framework for objective optic nerve (ON) assessment in vivo was successfully developed.
- This tool facilitates precise ON segmentation and quantification, advancing the study of related pathologies.
