This article is more than 5 years old. Information may no longer be current.
Beam positioning key to reliable retinal thickness measurements
Study finds great variance when the scanning beam was placed off axis from the center of the pupil.
Click Here to Manage Email Alerts
To obtain the most reliable retinal thickness measurements with spectral-domain optical coherence tomography, one study suggests that operators consistently position the scanning beam in the center of the pupil for a perpendicular angle between the OCT beam and retina rather than variably positioning it eccentrically at a non-perpendicular angle.
A study of 30 eyes of 15 healthy subjects with a mean age of 37.5 years found that the average increase in thickness for four non-perpendicular scan positions and subfields ranged from 3.76 µm to 11.38 µm.
The four eccentric positions evaluated were superior, inferior, nasal and temporal, and all were approximately 3 mm from the pupillary center.
SriniVas R. Sadda
There were only small differences in magnitude among the four eccentric positions, but scans displaced temporally showed the greatest increase in thickness and volume, while nasally positioned scans showed the least increase, according to study co-author SriniVas R. Sadda, MD.
“I don’t think we are prepared to make a conclusion that one eccentric position is necessarily worse than another,” Sadda said. “Still, you can obtain falsely elevated readings with any of these four eccentric positions.”
The study was published in Investigative Ophthalmology & Visual Science.
Measurement variability
Sadda manages an imaging and OCT reading center for clinical trials, the Doheny Image Reading Center, where assessing the reliability of results and considering factors that can confound, alter or otherwise affect the variability of measurements are of paramount importance, he said.
“We know that angle of incidence — the orientation or the direction at which the light beam used to image the retina actually strikes the retina — affects the visibility of certain structures. Therefore, we wanted to know whether it also causes a variability in the actual measurements,” Sadda told Ocular Surgery News. “Our hypothesis was that if you have an oblique angle of incidence as opposed to a perpendicular angle of incidence, you would expect the oblique incidence to cut through the retina in an oblique angle. Therefore, you would not be taking the most direct path through the retina, and that will translate to a greater thickness.”
Finding that the retinal thickness measurement increased when changing the angle of incidence perpendicularly was not surprising, Sadda said.
“The real interest was that I had no idea what the magnitude of retinal thickness would be. It was obviously greater than 10 µm,” he said, noting that 10 µm is often a threshold for meaningful difference in clinical trials.
The researchers made the same observations for retinal volume.
“Macular thickness and volume are absolutely related by area, and the area is constant,” Sadda said.
It is much easier to initially position the beam in the center of a dilated pupil and then repeat that same position the next time, Sadda said.
“A position off to the side is much harder to replicate so exactly the next time,” he said. “But we are less concerned about the repeatability of each individual eccentric position. Rather, if the operator takes the measurement one time in the center and then the next time is not careful and/or the patient is difficult to scan because he cannot keep his head still, the image may be taken slightly from the side, which could cause an increase in measurement.”
Clinical significance
For those designing clinical trials or in clinical practice, Sadda suggested that small apparent differences in retinal thickness over time may not be significant if it turns out the scan was obtained in a slightly different position.
Photographers may be content with taking a good image, regardless of the angle, unaware that it might falsely raise the measurement.
“We recommend that images be taken from the center every time,” Sadda said. “And if not, have the photographer make a note so that the clinician is aware.”
The study findings may benefit longitudinal studies, which require repeated measurements over time.
“For measuring fine or small changes, if you have a variable like retinal thickness, that could be a factor affecting your interpretation,” Sadda said. “You may falsely believe something is getting bigger or smaller when it is not because of small changes in position of how you scan the patient.”
Sadda said there is likely a way to correct for macular thickness variation through algorithms, based on measuring the angle of the scan displayed on the screen.
In the meantime, presentations and studies should state at what angle OCT images were obtained.
“The angle matters, and it matters to an extent that it is clinically relevant,” Sadda said. “If you measure from any direction aside from the exact center, the retinal thickness value measures greater. Regardless of what position you choose outside the center, those differences between eccentric positions are not likely significant. The fact that you are out of the center is what matters.” – by Bob Kronemyer