Benchmarking the Non-Paraxial Beam Propagation Method for Deep-Tissue Microscopy

This study evaluates the accuracy and computational efficiency of the Beam Propagation Method (BPM) for modeling light propagation in microscopy of biological tissue. Accurate simulation of light propagation is essential for improving imaging resolution and contrast in optical microscopy, where wavefront distortions and scattering pose major challenges. While rigorous solvers such as the Finite-Difference Time-Domain (FDTD) and Pseudo-Spectral Time-Domain (PSTD) methods provide high accuracy, their high computational cost makes large-volume simulations impractical for deep-tissue imaging. BPM extends the tractable depth of simulation assuming predominantly forward propagation in tissue, significantly reducing computational complexity. Unlike full-wave solvers, BPM avoids solving Maxwell’s equations in their entirety and requires fields to be stored for only two transverse planes at a time. This substantially lowers memory usage and computation time compared to full-field solvers, enabling the modeling of larger tissue volumes. We evaluate BPM’s performance using tissue models for multiphoton microscopy, where accurate modeling of wave propagation is critical. BPM is validated against Mie theory for single-sphere scattering and benchmarked against PSTD using experimentally acquired refractive index data and a discrete particle model that mimics biological tissue scattering. The Fast Fourier Transform BPM (FFT-BPM) is implemented without paraxial approximation, and the simulation results closely match reference solutions, while significantly reducing computational costs. These results establish BPM as a viable method for high-resolution optical microscopy in deterministically specified biological tissue. Its balance between accuracy and computational efficiency makes it particularly suitable for large-scale modeling. BPM’s extended simulation capabilities can advance biological imaging by improving our understanding of light-tissue interactions and enabling next-generation wavefront correction techniques.