Abstract
Purpose :
Overcome motion artifact during slow Brillouin acquisition by introducing 3-dimensional tracking. Demonstrate that motion-tracking Brillouin microscopy can distinguish keratoconus, post-LASIK or post-PRK from the control based on the different distribution of Brillouin shifts.
Methods :
Optical coherence tomography and pupil tracking were combined with a traditional Brillouin microscope to reduce motion blur caused by patient movement during slow Brillouin scan. Positioning errors were corrected in 3 dimensions by combing axial information from OCT and lateral information from pupil tracking. Tracking accuracy was tested in advance using enucleated porcine eyes. During in-vivo human measurement, 30 axial scans were applied to discrete points on each cornea. An axial scan took 5 seconds with a step size of 15 μm. 2-dimensional interpolation was used to connect the measured discrete points to generate a map of Brillouin shifts of the cornea.
Results :
Data from tests on porcine eyes showed that axial positioning errors were within 5 μm along a movement of 1 mm with a step size of 100 μm. Meanwhile, lateral positioning errors were within 3 μm over a 9 mm movement. In-vivo Brillouin results showed that a normal cornea, a subtle keratoconus, a post-LASIK and a post-PRK shared similar Brillouin shifts of around 5.7 GHz at the periphery. The difference mainly existed in the center region. The Brillouin shift at the center was about 5.69 GHz for the normal cornea, 5.67 GHz for the post-LASIK cornea and the post-PRK cornea. For the subtle keratoconus cornea, the Brillouin shift at the cone was about 5.65 GHz.
Conclusions :
Patient movement can be tracked precisely by pupil tracking and OCT. Brillouin shifts at different depths can be reaccommodated properly after motion compensation. Maps of Brillouin shifts across the cornea can show weak regions and help keratoconus diagnosis.
This abstract was presented at the 2022 ARVO Annual Meeting, held in Denver, CO, May 1-4, 2022, and virtually.