Demonstrate real-world applicability of predictive multi-landmark OCT tracking

Demonstrate the applicability of the predictive multi-landmark OCT tracking approach in a physical, real-world OCT tracking system beyond the simulated-motion evaluation.

Background

The paper introduces a predictive multi-landmark OCT tracking method that propagates positional updates between sequentially acquired landmarks by estimating a shared global pose change with the Kabsch algorithm. Simulated evaluations on porcine-skin OCT data show reliable tracking at velocities up to 100 mm/s for as many as nine landmarks.

The reported results are obtained under well-controlled simulated motion dynamics rather than during operation of the physical OCT tracking setup. Consequently, the method's applicability and performance in a real-world tracking system remain unresolved and require experimental demonstration on the physical system.