Optical coherence tomography distal-sensor guided manual trephine/dissection system for DALK

Abstract

Purpose : The most challenging aspect of deep anterior lamellar keratoplasty (DALK), is what’s known as “Big Bubble” technique which injects air/fluid to fully separate the Descemet’s Membrane (DM) and stroma with a hydro-dissection needle. Big bubble technique requires micron accuracy to guide the needle to approximately 90% depth of cornea. We hypothesize our OCT distal-sensor guided manual trephine/dissection system can guide the needle insertion by providing the position information of cornea layers with micron accuracy.

Methods : The manual DALK device uses suction to be fixed onto the eye. A second set of threads is used to advance the needle into the eye after trephination, at an angle of 60 degrees from vertical. A common-path swept source optical coherence tomography (CP-SSOCT) fiber sensor with a high index epoxy lens on the end surface was put inside a 27-gauge needle, and the sensor was inserted into the cornea stroma with the needle. The target depth of needle insertion was 95% of the cornea depth. The test was performed on human cadaveric eyes and we recorded A-scan images at different positions inside the cornea and its corresponding AB-mode images. We injected fluid to verify the results with the reference-based OCT system.

Results : The image in Fig. 1(a) shows the setup of our experiment. We have pre-measured the offset distance from the needle tip to the fiber reference plane which was 300 microns. Fig. 2 contains the A-scan images of each step and the corresponding AB-mode images. It should be noted that the cadaveric eyes exhibited swollen cornea with a thickness of approximately 800 microns. Fig.2(a) and (d) show the needle tip penetrating into the cornea where the distance between the needle tip and DM was approximately 800 microns. As the needle advanced forward as shown in Fig.2(b), 2(e) and 2(c), 2(f) the DM peak moves closer to the needle tip to the distance of 500 and 100 microns. We can verify the injected fluid was right above the DM within 100 microns from B-scan as Fig.1(c).

Conclusions : Our results are consistent with our hypothesis that manual DALK device can guide a needle insertion into corneal stroma for the big-bubble technique.

This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.

 

Fig1. (a) Experiment setup, (b) Image of human cornea after injection, (c) B-scan image after injection

 

Fig2. A-Scan images(left) and AB mode visualization (right) (a), (d)~800 microns, (b), (e) ~500 microns, (c), (f) ~100 microns from DM