Recollection, familiarity, and cortical reinstatement: a multivoxel pattern analysis - PubMed (original) (raw)

Recollection, familiarity, and cortical reinstatement: a multivoxel pattern analysis

Jeffrey D Johnson et al. Neuron. 2009.

Abstract

Episodic memory retrieval is thought to involve reinstatement of the neurocognitive processes engaged when an episode was encoded. Prior fMRI studies and computational models have suggested that reinstatement is limited to instances in which specific episodic details are recollected. We used multivoxel pattern-classification analyses of fMRI data to investigate how reinstatement is associated with different memory judgments, particularly those accompanied by recollection versus a feeling of familiarity (when recollection is absent). Classifiers were trained to distinguish between brain activity patterns associated with different encoding tasks and were subsequently applied to recognition-related fMRI data to determine the degree to which patterns were reinstated. Reinstatement was evident during both recollection- and familiarity-based judgments, providing clear evidence that reinstatement is not sufficient for eliciting a recollective experience. The findings are interpreted as support for a continuous, recollection-related neural signal that has been central to recent debate over the nature of recognition memory processes.

PubMed Disclaimer

Figures

Figure 1

Figure 1. Behavioral Performance

(A) Mean (+SEM) proportions of responses according to the test item condition. (B) Mean (+SEM) response time (RT) data. The Other category reflects collapsed Unsure Old, Unsure New, and Sure New responses (due to low individual trial numbers). The RT data for Remember responses to new items are based on only 12 subjects contributing such responses.

Figure 2

Figure 2. Importance Maps

Group mean importance maps for the three study tasks, overlaid on axial slices of the mean normalized anatomical data (coordinates in Talairach space). The colored areas depict voxels where importance values exceeded arbitrary thresholds of .001 positively (red) and -.001 negatively (green; see middle row, right-most column). L = left.

Figure 3

Figure 3. Classifier Accuracy

Mean classifier accuracy (+SEM) collapsed across all response categories and separated by response category. Time point (TR) 1 corresponds to test item onset. Shaded bars indicate the TRs during which classifier accuracy was significantly above chance (.33; correcting for multiple comparisons).

Figure 4

Figure 4. Classifier Output

Mean values (+SEM) of the classifier’s correct output node, (A) averaged over all three study tasks, and (B) over only the Artist and Function tasks. Each bar reflects classifier output for a given response category and time point (TR). Brackets indicate significant differences between responses (correcting for multiple comparisons).

Figure 5

Figure 5. Equivalent Reinstatement Effects

Results of searchlight analyses where reinstatement was equivalent for test items designated with Remember and Sure Old responses (see main text for details of the contrast procedure). Histograms reflect the mean (+SEM) output values at the correct classifier node (left column) and classifier accuracy (right column; chance = .33) within the depicted clusters in lateral temporal cortex, superior frontal gyrus, and inferior frontal gyrus. All effects depicted here survived a cluster-wise threshold of p < .05 and are overlaid on the mean anatomical image (coordinates in Talairach space). L = left.

Figure 6

Figure 6. Selective Reinstatement Effects

Results of searchlight analyses showing selective reinstatement for test items designated with Remember responses (compared to Sure Old responses; see main text for contrast procedure). The histograms provide the mean (+SEM) output value at the correct classifier node and the mean classifier accuracy within the depicted clusters of (A) posterior cingulate and (B) retrosplenial cortex. Both effects survived a cluster-wise threshold of p < .05. See Figure 5 caption for further display details.

Similar articles

Cited by

References

    1. Alvarez P, Squire LR. Memory consolidation and the medial temporal lobe: a simple network model. Proc. Natl. Acad. Sci. USA. 1994;91:7041–7045. - PMC - PubMed
    1. Bishop C. Neural Networks for Pattern Recognition. New York: Oxford University Press; 1995.
    1. Coltheart M. The MRC psycholinguistic database. Q. J. Exp. Psychol. A. 1981;33:497–505.
    1. Cox R. AFNI: software for analysis and visualization of functional magnetic resonance neuroimages. Comput. Biomed. Res. 1996;29:162–173. - PubMed
    1. Damasio AR. Time-locked multiregional retroactivation: a systems-level proposal for the neural substrates of recall and recognition. Cognition. 1989;33:25–62. - PubMed

Publication types

MeSH terms

LinkOut - more resources