Using structural aberration coefficients to filter first-order starting points for reflective freeform zoom ophthalmoscopes

Adaptive optics ophthalmoscopes are optical systems used to image retinas in vivo at single-cell resolution. These systems have traditionally been designed with a fixed exit-pupil size to match a typical dilated pupil size, but an optical zoom system introduces the flexibility to adapt for a wider patient range. Reflective freeform zoom ophthalmoscope systems are complex designs whose performance depends heavily on a well-defined optimization process. A common design approach is to generate a large number of first-order starting points, which are paraxial layouts that are candidates for further optimization. This provides a way to search the broad solution space, but currently there is no means of filtering through the numerous starting points other than optimizing each starting point individually in optical design software. This is a time consuming and inefficient optimization process, so a filtering method for the starting points is desirable. We first investigate a paraxial, in-line (rotationally symmetric) optimization method that minimizes the structural aberration coefficients for spherical aberration, coma, and astigmatism across zoom positions. Then we show that the resulting in-line error metric correlates to the final, unobscured freeform performance of the system, providing a metric to rank each system’s potential earlier in the design process. This eliminates the optimization of each starting point in design software, making the design process much more efficient.