Data-driven optimization of preschoolers’ hemodynamic response in a VR setup: advancing analytic methods for children’s fNIRS naturalistic data with the AICopt method

Figures

• 

  Fig. 1. Cave automatic virtual environment (CAVE) at the Birkbeck ToddlerLab, and equipment worn by children during the experiment.

• 

  Fig. 2. Depiction of the scene seen by the toddlers for an example of ER and NER scenarios.

• 

  Fig. 3. (a) Representation of the channels on a 5-year-old multilayer head model mesh; (b) schematic representation of the channels overlapped onto the gray matter mesh extracted from the multilayer volumetric template model. ROIs are highlighted: orange represents MPFC, red represents DLPFC, green represents M/STG, blue represents TPJ, yellow represents IPL. Figure reproduced with permission from Bulgarelli et al.5

• 

  Fig. 4. Schematic summary of the main steps of the AICopt workflow.

• 

  Fig. 5. Group-averaged HRs for HbO2 (in red) and HHb (in blue), comparing the ER condition (solid line) with the NER condition (dotted line) between 2 s before the event stimuli (yellow shaded area) onset and 20 s past the action block (green shaded area) start.

• 

  Fig. 6. Distribution of peak response amplitudes for both HbO2 (a) and HHb (b) across ROIs. Distributions marked with an asterisk are the ones for which the null hypothesis was not rejected.

• 

  Fig. 7. T-value maps derived for the ER event versus NER event contrast using the block averaging pipeline. Positive t-values indicate increased concentration for both HbO2 and HHb. The expected task-related activation consists of increased HbO2 and decreased HHb concentrations. Regions showing uncorrected statistical significance are outlined with a white contour, whereas those surviving FDR correction for multiple comparisons are marked with an asterisk.

• 

  Fig. 8. T-value maps derived for the ER event versus NER event contrast using the canonical GLM pipeline. Positive t-values indicate increased concentration for both HbO2 and HHb. The expected task-related activation consists of increased HbO2 and decreased HHb concentrations. Regions showing uncorrected statistical significance are outlined with a white contour.

• 

  Fig. 9. T-values maps derived for the ER choice versus NER choice contrast using the canonical GLM pipeline. Positive t-values indicate increased concentration for both HbO2 and HHb. The expected task-related activation consists of increased HbO2 and decreased HHb concentrations. Regions showing uncorrected statistical significance are outlined with a white contour, whereas those surviving FDR correction for multiple comparisons are marked with an asterisk.

• 

  Fig. 10. Average HRF models estimated using the AICopt method for the ER (in blue) and NER (in yellow) conditions. The canonical HRF model is also shown in black. The shaded areas around the average curves represent the standard deviation across the considered ROIs.

• 

  Fig. 11. T-value maps derived for the ER event versus NER event contrast using the AICopt-based GLM pipeline. Positive t-values indicate increased concentration for both HbO2 and HHb. The expected task-related activation consists of increased HbO2 and decreased HHb concentrations. Regions showing uncorrected statistical significance are outlined with a white contour, whereas those surviving FDR correction for multiple comparisons are marked with an asterisk.

• 

  Fig. 12. T-value maps derived for the ER choice versus NER choice contrast using the AICopt-based GLM pipeline. Positive t-values indicate increased concentration for both HbO2 and HHb. The expected task-related activation consists of increased HbO2 and decreased HHb concentrations. Regions showing uncorrected statistical significance are outlined with a white contour.