Figure 1.
(a) Experimental stimuli. Sample illustrations of the 50 familiar objects (FO) and 100 novel non-objects (LNO, UNO) used in this study. Functional properties were taught for half of the 100 non-objects during a behavioral learning session after MEG session 1 (S1), thus becoming learned non object (LNO) at session 2 (S2). No associations were taught for the other 50 non-objects that remained unknown (unlearned) non-objects (UNO) at both sessions 1 and 2. (b) MEG picture naming/defining task: children were asked to provide the name and/or a definition of the object presented on the screen after the appearance of the question mark 1 sec after stimulus presentation, or to signal if the object was unknown.
Table 1.
Modulations of brain responses (ERFs) in the sensor space, associated with the learning of novel objects properties (LNO S1<S2).
Figure 2.
Learning-related changes in evoked-related fields (ERFs) over the right temporal region (analysis in sensor space).
Two dimensional statistical map (left, top side) showing sensor-space differences (t-test, p<.001) in field intensities of LNO evoked-responses between sessions (LNO S1< S2), at 740 msec post-stimulus onset masked by the repetition effect for UNO objects (S1 vs. S2). Red points indicate the location of MEG gradiometers (26 33 and 26 32) eliciting evoked responses represented on Figure 2 (a, b and c). (a) Grand average time courses of ERFs for LNO and UNO non-objects at S1 (LNO: black hyphenated line, UNO: gray hyphenated line) and S2 (LNO: black line; UNO: gray line). Significant differences are identified 720–740 msec post-stimulus onset. (b) Grand average time course of ERFs for FO (light gray hyphenated line), pre-learning LNO (black hyphenated line) and UNO (gray hyphenated line) at S1. (c) Grand average time course of ERFs for FO (light gray line), pre-learning LNO (black line) and UNO (gray line) at S2.
Figure 3.
Learning the Functions of Novel Objects: Brain Sources.
Top panels: Source reconstruction of neural activity elicited during picture definition in each session separately for LNO objects during significant time windows identified in the sensor space (a) 530–690 ms, (b) 700–760 ms (left subpanel, LNO S1 pre-learning session; right subpanel, LNO S2 post learning session; displayed at p<.001 uncorrected). Colors code the value of the F statistic associated with each voxel. Bottom panels: source reconstructions for neural activity underlying learning-related effects (LNO S1<S2: P<.005 uncorrected) during the same time windows: (a) bilateral activations over the orbito-frontal gyri and the right parietal posterior gyrus at 530–690 msec; (b) right lateral occipito-temporal and infero-temporal activations at 700–760 msec. Colors code the value of the t statistic associated with each voxel.
Table 2.
Modulation of brain activity in the source space, associated with the learning of novel objects properties (LNO S1<S2), masked exclusively for between-sessions repetition effects for untaught non objects (exclusive mask puncorr <.05; UNO S1 vs. S2).