Improved Automatic Optic Nerve Radius Estimation from High Resolution MRI.

Area of Science:

• Neuroimaging
• Ophthalmology
• Medical Image Analysis

Background:

• The optic nerve (ON) is crucial for vision, transmitting retinal information to the brain.
• Current methods for measuring ON damage lack precision, assessing only arbitrary points.
• Optic nerve damage correlates with visual deficits due to the visual pathway's lack of redundancy.

Purpose of the Study:

• To develop and validate a fully automated, 3D-consistent MRI technique for estimating optic nerve and cerebrospinal fluid radius.
• To improve the accuracy and reproducibility of optic nerve measurements compared to slice-wise methods.
• To assess the long-term reproducibility of the new technique for detecting subtle anatomical changes.

Main Methods:

• A novel, fully automated, 3D-consistent technique was developed based on a prior slice-wise method.
• The technique estimates optic nerve and cerebrospinal fluid radius from high-resolution, T2-weighted isotropic MRI scans.
• The new method was compared against the previous slice-wise technique using short-term (<1 month) and long-term (approx. 11 months) reproducibility datasets.

Main Results:

• The 3D-consistent technique produced more anatomically viable results than the slice-wise method.
• The new method demonstrated superior reproducibility, particularly in the central optic nerve region, which is less affected by motion and interference.
• Long-term reproducibility was similar to short-term reproducibility, suggesting optic nerve size is stable over approximately 11 months.

Conclusions:

• The developed 3D-consistent MRI technique offers enhanced accuracy and reproducibility for optic nerve measurements.
• This improved methodology increases the power to detect anatomical changes in optic nerve size within patient populations.
• The findings support the stability of optic nerve anatomy over an 11-month period.