Research ArticleNEUROSCIENCE

Does the mismatch negativity operate on a consciously accessible memory trace?

See allHide authors and affiliations

Science Advances  13 Nov 2015:
Vol. 1, no. 10, e1500677
DOI: 10.1126/sciadv.1500677
  • Fig. 1 Stimulation paradigm and behavioral results.

    (A) Spectrographic representation of a targets-present masked trial (left), a targets-absent masked trial (middle), and an unmasked control trial (right). (B) True- and false-positive rates (that is, hits and false alarms) as a function of position within the stimulus sequence. (C) Corresponding d′. (D) True-positive rates as a function of position, binned by target frequency.

  • Fig. 2 Neural activity in the control (A to C) and masked (D to F) conditions.

    (A) Grand-average normalized dSPM between 100 and 150 ms for deviants minus standards under control (unmasked) conditions. (B) Grand-average MxNE solutions for (top) standards (black traces) and deviants (gray traces) and (bottom) their respective difference waveforms. (C) Quantified amplitudes for standards (black) and deviants (gray). (D) Grand-average MxNE solutions for virtual targets (gray), deviants and standards for undetected standard streams (blue), deviants and standards for detected standard streams (orange), and their respective difference waveforms (deviants minus standards). (E) Corresponding dipole locations. (F) Quantified amplitudes for MMN responses generated in the context of detected (orange) and undetected (blue) standard streams. au, arbitrary units.

  • Fig. 3 Neural activity under masked conditions, with emphasis on the P1 latency range.

    (A) Grand-average MxNE solutions for virtual targets (gray), deviants and standards for undetected standard streams (blue), deviants and standards for detected standard streams (orange), and their respective difference waveforms. (B) Corresponding dipole locations. (C) Quantified amplitudes for P1 responses generated in the context of detected (orange) and undetected (blue) standard streams, for both standards (left) and deviants (right).

  • Fig. 4 Neural activity elicited by masker tones, with emphasis on the P1 latency range.

    (A) Grand-average MxNE solutions for responses elicited by masker tones for targets-detected trials (solid traces; top) and targets-undetected/targets-absent trials (dotted traces; bottom) for early (green) and late (magenta) time intervals. Epochs for targets-detected trials were created by splitting those trials into two time intervals (before and after target detection). Epochs for targets-undetected/targets-absent trials were created by splitting those trials into two time periods based on the average detection time of the targets-detected trials. (B) Corresponding dipole locations. (C) Quantified P1 amplitudes in response to masker tones both before (green) and after (purple) detection of the standard stream. Closed (open) bars are for targets-detected (targets-undetected) trials. A three-way ANOVA with hemisphere, target detection, and time interval as factors revealed significant two-way interactions between hemisphere and interval (F1,19 = 9.2, P < 0.01) as well as a marginally significant two-way interaction between target detection and interval (F1,19 = 3.3, P = 0.09). However, subsequent paired comparisons did not show significant effects of the time interval on either left AC or right AC in the direction one would expect based on an attentional account of the data (table S2). A similar pattern of results was observed when the more anterior MMN-defined source space was used (cf. figs. S4 and S6).

  • Fig. 5 Neural activity elicited by masker tones, with emphasis on the N1 latency range.

    (A) Grand-average MxNE solutions. (B) Corresponding dipole locations. (C) Quantified N1 amplitudes. A three-way ANOVA with hemisphere, target detection, and time interval as factors revealed a significant main effect of interval (F1,19 = 15.1, P < 0.005) as well as a significant two-way interaction between interval and target detection (F1,19 = 5.1, P < 0.05). However, this interaction went in the opposite direction that would be expected based on an attentional account of the data (that is, the early-late difference was larger for targets-undetected and targets-absent trials; table S2). Here, the source space used for the masker-elicited N1 was the same as for the masker-elicited P1. The same pattern of results was found when the more anterior MMN-defined source space was used (cf. figs. S5 and S6).

Supplementary Materials

  • Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/1/10/e1500677/DC1

    Fig. S1. Neural activity under masked conditions after equalization of the number of epochs in each frequency bin across detected and undetected standards and deviants.

    Fig. S2. Neural activity under control conditions, binned by frequency.

    Fig. S3. Neural activity under masked conditions after epoch equalization, with emphasis on the P1 latency range.

    Fig. S4. Masker-elicited P1 responses using the same source space as was used for MMN.

    Fig. S5. Masker-elicited N1 responses using the same source space as was used for MMN.

    Fig. S6. Source-space vertices for P1 and MMN source analyses.

    Table S1. Target-tone P1 statistics.

    Table S2. Masker-tone P1/N1 statistics.

    Audio S1. Example of an auditory oddball sequence (an isochronous 902-Hz standard stream with a 947-Hz deviant and a 500-ms stimulus onset asynchrony) embedded in a random multitone masker cloud.

    Audio S2. Example of an auditory oddball sequence in isolation.

  • Supplementary Materials

    This PDF file includes:

    • Fig. S1. Neural activity under masked conditions after equalization of the number of epochs in each frequency bin across detected and undetected standards and deviants.
    • Fig. S2. Neural activity under control conditions, binned by frequency.
    • Fig. S3. Neural activity under masked conditions after epoch equalization, with emphasis on the P1 latency range.
    • Fig. S4. Masker-elicited P1 responses using the same source space as was used for MMN.
    • Fig. S5. Masker-elicited N1 responses using the same source space as was used for MMN.
    • Fig. S6. Source-space vertices for P1 and MMN source analyses.
    • Table S1. Target-tone P1 statistics.
    • Table S2. Masker-tone P1/N1 statistics.
    • Legends for audio S1 and S2

    Download PDF

    Other Supplementary Material for this manuscript includes the following:

    • Audio S1 (.wav format). Example of an auditory oddball sequence (an isochronous 902-Hz standard stream with a 947-Hz deviant and a 500-ms stimulus onset asynchrony) embedded in a random multitone masker cloud.
    • Audio S2 (.wav format). Example of an auditory oddball sequence in isolation.

    Files in this Data Supplement:

Navigate This Article