Data-efficient extraction of optical properties from 3D Monte Carlo TPSFs using Bi-LSTM transfer learning

TL;DR

This paper introduces a physics-informed transfer learning approach using Bi-LSTM networks to efficiently extract optical properties from 3D Monte Carlo measurements in real-time, reducing bias and computational cost.

Contribution

It presents a novel transfer learning strategy that combines deterministic solvers and stochastic simulations for rapid, accurate optical property estimation.

Findings

Eliminates systematic bias of analytical models

Achieves near-instantaneous inference

Bridges analytical and stochastic domain gap

Abstract

Time-Resolved Spectroscopy (TRS) is a powerful modality for non-invasive characterization of turbid media. However, extracting optical properties, absorption $\mu_a$ and reduced scattering $\mu_s'$, from 3D stochastic measurements remains computationally expensive for real-time applications. In this paper, we propose a data-efficient, physics-informed transfer learning strategy using a Bidirectional Long Short-Term Memory (Bi-LSTM) network. By leveraging a fast deterministic solver to establish a physical prior before fine-tuning on a restricted set of 3D Monte Carlo simulations, our model successfully bridges the analytical-to-stochastic domain gap. The proposed method eliminates the systematic bias of analytical models while maintaining a competitive error with near-instantaneous inference time.