Phase function estimation from a diffuse optical image via deep learning

TL;DR

This paper introduces a deep learning approach to estimate the phase function from diffuse optical images without assuming a specific form, using a Gaussian mixture model to improve Monte Carlo light simulation accuracy.

Contribution

It develops a convolutional neural network that estimates the phase function directly from images, generalizing beyond traditional parametric models like Henyey-Greenstein.

Findings

Mean squared error of phase function estimation is 0.01

Relative error of anisotropy factor is 3.28%

Method validated on simulated biological tissue images

Abstract

The phase function is a key element of a light propagation model for Monte Carlo (MC) simulation, which is usually fitted with an analytic function with associated parameters. In recent years, machine learning methods were reported to estimate the parameters of the phase function of a particular form such as the Henyey-Greenstein phase function but, to our knowledge, no studies have been performed to determine the form of the phase function. Here we design a convolutional neural network to estimate the phase function from a diffuse optical image without any explicit assumption on the form of the phase function. Specifically, we use a Gaussian mixture model as an example to represent the phase function generally and learn the model parameters accurately. The Gaussian mixture model is selected because it provides the analytic expression of phase function to facilitate deflection angle sampling in MC simulation, and does not significantly increase the number of free parameters. Our proposed method is validated on MC-simulated reflectance images of typical biological tissues using the Henyey-Greenstein phase function with different anisotropy factors. The effects of field of view (FOV) and spatial resolution on the errors are analyzed to optimize the estimation method. The mean squared error of the phase function is 0.01 and the relative error of the anisotropy factor is 3.28%.