Abstract
Purpose :
The optic nerve is a communication pathway between the brain and the eye. The subarachnoid space between the nerve and sheath contains cerebrospinal fluid that expands with increased intracranial pressure. US-determined ONSD can be a proxy for intracranial pressure but methods suffer from limited reproducibility due to the nerve’s coaxial presentation with respect to the US beam axis in primary gaze and its small diameter. Measurement of ONSD is time-consuming and subject to high variability. In this study, we compare ONSD measurements derived from a mechanically scanned single-element probe with a linear array.
Methods :
We scanned both eyes of 10 subjects with a 15 MHz single-element probe (Quantel Absolu) and a linear array of comparable frequency (GE Venue Go with L4-20t-RS probe). Scans were performed in a horizontal plane. ONSD was measured 3 mm posterior to the optic nerve head in 6 images selected from 3 cineloops. Measurements within and between two investigators were compared by Bland Altman analysis.
Results :
SDs within 6 measurements of the same eye were greater on the Absolu (0.44 mm) than the Venue (0.22 mm). ONSD values were slightly higher on the Absolu than the Venue by 0.22±0.45 mm by user 1 and 0.10±0.47 mm by user 2. Correlation coefficients(R) between the instruments were 0.74 and 0.57 for users 1 and 2. R between users was 0.73 on the Absolu, 0.79 on the Venue and 0.88 on the mean value of Absolu and Venue determinations .
Conclusions :
Patients suspected of having optic disc edema are often subjected to invasive procedures which could be avoided using an alternative imaging modality such as US measurement of OSND. US image quality is related to probe focal properties and bandwidth. Single-element probes have a symmetrical beam with a fixed focus, usually just anterior to the optic nerve head. Linear arrays typically beamform to maintain a fixed F-ratio with depth, but have poor out-of-plane focusing. We found the linear array to provide less speckle than the single-element probe. The nerve sheath was more clearly defined because of this, resulting in reduced variation among measurements. We conclude that a broadband high frequency linear array is superior for this application than a typical B-scan probe.
This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.