Models for light propagation in the head: trends and implementation of Finite Element approaches

• DOI: 10.1016/j.brs.2024.12.975
• Publication Date: 2025-02-25T05:35:27Z
• Authors: Filipa Fernandes; Filipe S. Silva; Nuno Sousa; Oscar Carvalho; Susana O. Catarino
• Keywords: Saúde de qualidade; PBM; Brain; Ciências Médicas::Biotecnologia Médica; Neurosciences. Biological psychiatry. Neuropsychiatry; Numerical simulation; Finite Element; RC321-571

The simulation of light propagation in the head tissues has become of interest due to its applicability in transcranial photobiomodulation (tPBM). In silico studies appear as an effective way to research tPBM dosimetry and understand the light reach in the head tissues, as well as potential issues regarding tissue heating. Nonetheless, these computer models are still very simple and come with several limitations. Although Monte Carlo methods are still the most common models for the simulation of light propagation in the human head, deterministic models – mostly based on Finite Element Methods (FEM) – are becoming increasingly important and could offer a more computationally efficient solution for optical propagation modelling. Regarding tissues’ optical simulation, literature shows that the scalp, skull and brain are the most studied tissues, sometimes including the distinction between white and grey matter. These models show limitations, namely the complex geometry of the brain surface, which is usually not considered; and the oversight of the cerebrospinal fluid (CSF), which due to its low/non-scattering properties cannot be simulated using the diffusion approximation equation. One improvement that can be considered for future research is the standardization of the reduced scattering and absorption coefficients of the tissues, as studies report varying values. With this information, the skin, cranium and brain were simulated using FEM in COMSOL Multiphysics software. The scattering and absorption coefficients of the tissues were derived from values reported in literature. The simulations produced results closely aligned with the reviewed data, confirming that the selected parameters and approaches are adequate for reproducing expected outcomes in tPBM.