Simultaneous integrated diffuse optical tomography and functional magnetic resonance imaging of the human brain

Abstract

A complete methodology has been developed to integrate simultaneous diffuse optical tomography (DOT) and functional magnetic resonance imaging (MRI) measurements. This includes development of an MRI-compatible optical probe and a method for accurate estimation of the positions of the source and detector optodes in the presence of subject-specific geometric deformations of the optical probe. Subject-specific head models are generated by segmentation of structural MR images. DOT image reconstruction involves solution of the forward problem of light transport in the head using Monte Carlo simulations, and inversion of the linearized problem for small perturbations of the absorption coefficient. Initial results show good co-localization between the DOT images of changes in oxy- and deoxyhemoglobin concentration and functional MRI data.

Figures

Fig. 1

(a) Schematic of the fMRI-compatible optical probe. (b) Optode positions reconstructed from MR images, superimposed on a surface rendering of a 3D MRI data set.

Fig. 2

a) Simulated activation (modeled as a change in absorption coefficient) compared with the (b) reconstructed result. One-dimensional projections of the simulated and reconstructed absorption perturbations are plotted as functions of distance from the center of the cross-hairs along the X (c) and Y (d) coordinates.

Fig. 3

(a–c) Spatial maps of the correlation coefficient, shown as the colour bar on the right. (a) Spatial hemodynamic response derived from the BOLD MRI signal, (b) spatial hemodynamic response corresponding to the change in deoxyhemoglobin concentration derived from the reconstructed optical data, (c) spatial hemodynamic response corresponding to the change in oxyhemoglobin concentration derived from the reconstructed optical data. (d–e) Temporal changes in oxy- and deoxyhemoglobin concentrations derived from the optical data. (d) Temporal hemodynamic response of corresponding to an average of all voxels with a correlation coefficient greater than 0.5, (e) response from voxels within the sensitivity region but with a correlation coefficient less than 0.5, and (f) response from the single voxel showing the maximum changes in the oxy- and deoxyhemoglobin concentrations.