Diffuse optical reconstructions of fNIRS data using Maximum Entropy on the Mean

Abstract

Functional near-infrared spectroscopy (fNIRS) measures the hemoglobin concentration changes associated with neuronal activity. Diffuse optical tomography (DOT) consists in reconstructing the optical density changes measured from scalp channels to the near-infrared light attenuation changes within the cortical regions. In the present study, we adapted a nonlinear source localization method developed and validated in the context of Electro- and Magneto-Encephalography (EEG/MEG): the Maximum Entropy on the Mean (MEM), to solve the inverse problem of DOT reconstruction. We first introduced depth weighting strategy within the MEM framework for DOT reconstruction to avoid biasing the reconstruction results of DOT towards superficial regions. We also proposed a new initialization of the MEM model improving the temporal accuracy of the original MEM framework. To evaluate MEM performance and compare with widely used depth weighted Minimum Norm Estimate (MNE) inverse solution, we applied a realistic simulation scheme which contained 4000 simulations generated by 250 different seeds at differ- ent locations and 4 spatial extents ranging from 3 to 40cm2 along the cortical surface. Our results showed that overall MEM provided more accurate DOT reconstructions than MNE. Moreover, we found that MEM was remained particularly robust in low signal-to-noise ratio (SNR) conditions. The proposed method was further illustrated, by comparing to functional Magnetic Resonance Imaging (fMRI) activation maps, on real data involving finger tapping tasks with two different montages. The results showed that MEM provided more accurate HbO and HbR reconstructions in spatial agreement with the fMRI main cluster, when compared to MNE.

Evaluation of a personalized functional near infra‐red optical tomography workflow using maximum entropy on the mean

“…We compared two fNIRS reconstruction methods, one being our previously proposed and validated maximum entropy on the mean (MEM) (Cai et al, 2021) and the other one being the conventional minimum norm estimation (MNE) (Hämäläinen & Ilmoniemi, 1994).…”

“…This observation is due to the limited ability of fNIRS to detect the deeper areas, a standard limitation of the technique itself (Scholkmann et al, 2014) which can be partially compensated by introducing depth weighting when solving the inverse problem. We investigated the impact of depth weighting, in both MEM and MNE, in fNIRS reconstruction in our previous study (Cai et al, 2021). The optimal depth weighting parameters identified in our previous study were actually the ones considered for this study.…”

“…A key property of MEM source imaging is its ability to accurately recover the spatial extent of the generators, as demonstrated in the context of: (a) localizing transient epileptic discharges (Chowdhury et al, 2016; Grova et al, 2016; Heers et al, 2016; Pellegrino et al, 2016; Pellegrino et al, 2020) and oscillations (Avigdor et al, 2020; Pellegrino, Hedrich, et al, 2016); (b) localizing focal sources, such as those evoked by electrical median nerve stimulations (Hedrich, Pellegrino, Kobayashi, Lina, & Grova, 2017); (c) EEG/MEG fusion in the presurgical evaluation of epilepsy (Chowdhury et al, 2018); and (d) MEG resting state connectivity (Aydin et al, 2020). In our previous study (Cai et al, 2021), we adapted the MEM framework to perform fNIRS reconstructions to generate NIROT images and then carefully evaluated MEM performance within a comprehensive and realistic simulation framework. In this study, we opted to combine the above methodology developments as a workflow for conducting NIROT and evaluated its performance using the real data acquired during a motor task.…”

Hemodynamic Correlates of Fluctuations in Neuronal Excitability: A Simultaneous Paired Associative Stimulation (PAS) and functional Near Infra-Red Spectroscopy (fNIRS) Study

“…The details of the following procedures are provided in Appendix B. Briefly, fNIRS data processing was performed applying 3D reconstructions with the Maximum Entropy on the Mean (MEM) framework, as described in [Cai et al, 2021a;Cai et al, 2021b]. The goal was to extract HbO/HbR amplitude from spatiotemporal maps reconstructed from the channel space data to the underlying cortical surface.…”

Hierarchical Bayesian Modeling of the Relationship between Task Related Hemodynamic Responses and Cortical Excitability

“…Meanwhile, the MEM framework adapted from our previous works in the context of electro-/magneto-encephalogram source imaging (Chowdhury et al, 2013(Chowdhury et al, , 2016Grova et al, 2016;Heers et al, 2016;Hedrich et al, 2017;Pellegrino et al, 2020) for conducting fNIRS reconstruction (Cai et al, 2021a) also ensured accurate estimation of HbO/HbR time courses from reconstructed spatiotemporal maps. For instance, delays between HbO and HbR peak times were around 1s shown in Fig.…”