Skip to main content
. 2019 Jul 10;11:12. doi: 10.1186/s11689-019-9272-2

Fig. 3.

Fig. 3

To extract information of affective nature, three regions of interest (ROIs) were used (tip of the nose, forehead, and cheeks). A mask was initially created (a) in which the ROIs were drawn to ensure that there was no variability across the size and shape of the ROIs among participants. Taking as a reference point the nose, we traced an imaginary straight line centered on the tip of the nose to locate the area of the forehead placed above the two eyebrows. Further horizontal lines passing through the center of the eyes and the tip of the nose allowed us to identify the area of the cheeks. The same mask was uploaded on each participant’s first frame and then a trained experimenter analyzed frame by frame the movements of the participant’s face during the experiment. Whenever the participant moved, the experimenter adjusted the position of each ROI on the participant’s face. If the participant’s movements did not allow the reposition of the ROIs, the frame was discarded. In b, we reported two example of thermal extraction (in the top panel a child with Moebius syndrome and in the lower panel a child belonging to the control group). In the images of participants, blurred masks (glasses) are included on the children’s face in order to occlude their identity