Optical coherence tomography (OCT) of the rim and peripapillary retina can aid in the diagnosis of glaucoma and its progression.
The Avanti (Optovue) also acquires a grid scan, and rim measurements are made relative to the Bruch’s membrane opening. The Avanti reports the ONH parameters in a manner similar to the other machines, but it also provides the angiography data (AngioAnalytics), which is the only instrument to date that has received FDA approval for both structure and angiography.
The Swept Source (Topcon) also acquires grids, the sizes of which are based on operator preference. Disc detection points are determined based on the edge points of the retinal pigment epithelium, rather than using the Bruch’s membrane opening, to define the disc and establish a base plane, and a circle scan is extracted from the A-scan grid.
However, he explained, the rim width and cup volume are measured relative to or within a reference plane above the base plane, which is similar to how measurements were done in confocal scanning laser tomography. The Swept Source includes a normative database, which is pending FDA approval.
Dr. Burgoyne pointed out that the findings in the current literature regarding glaucoma detection are “very encouraging.”
“We are now achieving greater than 95% specificity and detecting glaucoma (defined by the presence of glaucomatous visual field loss) with from 80% to 95% sensitivity, depending on the severity of the visual field loss,” he said. “Moderate to advanced disease, i.e., that which exceeds -4 to -6 decibels is detected with greater than 95% sensitivity.”
“Both the minimal rim width and peripapillary RNFL thickness are being used and their performance is similar,” he explained. “However, using them together fosters the eventual integration of macular data. OCT angiography may be helpful, but is still early in development.”
Dr. Burgoyne predicted that ONH anatomy will be integrated with the nerve fiber layer and the macula to increase detection of early glaucoma.
“This will help clinicians improve their examination,” he concluded. “Deep imaging targets will emerge on the instruments in the near future. Deep learning is being applied to improve the performance of the current algorithms that will open a new phase of development.”
Claude F. Burgoyne, MD
E: [email protected]
Dr. Burgoyne receives unrestricted research support from and is an unpaid consultant to Heidelberg Engineering. He is also NIH 401 funded to study the issues presented in this talk. Dr. Burgoyne disclosed that the information about each instrument was provided by the specific manufacturer and should be confirmed by each.