Abstract
Purpose :
To estimate the sensitivity of PhD-OCT measurements at distinguishing various disease, treatment, and healing states of the cornea. To date, 203 eyes have been assessed (7 of which have manifest keratoconus and 36 of which were eyes imaged both pre- and post- refractive surgery).
Methods :
Patients were imaged with a commercial anterior segment scanner (Bioptigen Envisu C-class). The imaging probe was mounted to a slit-lamp stand and the 12-mm telecentric objective lens for anterior imaging was used. Both M-B mode scans and B-mode scans were acquired. A simplified decorrelation metric is calculated for each 6px axial window in the M-scan. Axial motion in the M-scan is compensated by correcting the phase shift due to the Doppler effect. To compute lateral averages, the corneal surface in each M-scan is computationally shifted to a fixed offset. (Fig 1) Follow-up images were aquired approxiamtely 3 months post-surgery.
Results :
There is a clear difference between pre- and post- surgical eyes, with the post- eyes exhibiting higher decorrelation, corresponding to micromechanical weakening. This difference is apparent when all eyes are considered (Fig 2) as well as when individual eyes are compared before and after surgery. These differences are not readily observed with structural image characteristics such as intensity or speckle contrast. The mean p-value of averaged corneal profiles (pre- vs post-, by eye) is p=0.0017.
Conclusions :
The higher decorrelation values exhibited by post-surgical corneas indicate that micromechanical weakening occurs as a result of surgery. Further investigation and patient imaging will determine if the magnitude of the micro-mechanical change is associated with the total volume of tissue removed or with other recovery metrics. Because this technique uses conventional OCT equipment, it could provide a readily accessible method for assessing the corneal response to surgery and identifying eyes with elevated risk for complications such as corneal ectasia.
This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.