Article Figures & Data
Figures
- Figure 1.
Working memory paradigm. After a pair of tones was presented, subjects were cued to maintain one of the two tones in working memory for 16 s. A decision period followed in which subjects compared a probe tone with the tone in working memory.
- Figure 2.
Activity in the auditory cortex during encoding, maintenance, and retrieval phases of the working memory task. A, Results of univariate analyses at the individual voxel level for encoding (top), maintenance (middle), and retrieval (bottom) phases. Activity is thresholded at p = 0.05, corrected for multiple comparisons for the volume of auditory cortex. B, Comparison of activity in different ROIs of the auditory cortex for encoding (top), maintenance (middle), and retrieval (bottom) phases. C, Time series from different ROIs of auditory cortex. Med, Medial; Mid, middle; Lat, lateral.
- Figure 3.
Activity in hippocampus during encoding, maintenance, and retrieval phases of the working memory task. A, Results of univariate analysis at the individual voxel level for encoding (top), maintenance (middle), and retrieval (bottom) phases. Activity is thresholded at p = 0.05 corrected for multiple comparisons for the volume of hippocampus. B, Activity for the three phases of the working memory task along the anterior–posterior axis of left hippocampus (top) and right hippocampus (bottom). This is done by calculating the average activity of all voxels within hippocampus at a given value of y-coordinate. The y-coordinate is changed from anterior to posterior (y = −8 to −38) at 2 mm resolution. C, Time series plot of activity in the anterior and posterior hippocampus in left hemisphere (i) and right hemisphere (ii).
- Figure 4.
Activity in the left and right IFG for the encoding, maintenance, and retrieval phases of the working memory task. Ai, Results of univariate analysis at the individual voxel level overlayed on the structural scan. Peaks of activity in the left and right IFG are marked with white circles. Activity is thresholded at p = 0.05 corrected for multiple comparisons for the whole brain. Aii, Plots of β values at the peak voxels in left and right IFG during the encoding, maintenance, and retrieval phases. Aiii, Time series plots at the peak voxels of left and right IFG. Bi, A more dorsal–anterior (than shown in Ai) location in the right IFG that only responds during the retrieval phase. Plots of β values (Bii) and time series (Biii) at that locus are shown.
- Figure 5.
A–C, Correlation of BOLD activity with behavioral performance accuracy during the encoding (A), maintenance (B), and retrieval (C) phases. Voxels with significant correlation with behavioral performance were identified using p = 0.001 (uncorrected). Ai, Aii, Positive correlation during encoding is observed in rTPJ (Ai) and STG/STS (Aii). Bi–Biv, Positive correlation during maintenance in parietal cortex (Bi); negative correlation during maintenance in DMN (Bii), hippocampus (Biii), and IFJ (Biv). Ci, Negative correlation during retrieval in retrosplenial cortex. For display purposes, β values overlaid in the MRI structural scan during encoding, maintenance and retrieval are thresholded at p = 0.01 (uncorrected), p = 0.001 (uncorrected) and p = 0.001 (uncorrected) respectively.
- Figure 6.
Percentage accuracy of pattern classification (low vs high tone) during the maintenance period using MVPA. The mean (±SEM across subjects) classifier performance is shown. Chance performance is 50% (marked horizontal line). Accuracy is above chance level in the HG (p = 0.001) and left IFG (p = 0.016).
- Figure 7.
Analysis of connectivity among auditory cortex, IFG, and hippocampus. Connectivity is compared during the three phases of the working memory task. i, Encoding vs maintenance, p = 0.05 (corrected). ii, Retrieval > encoding (p = 0.05 (corrected). No connection was significant for encoding > retrieval. There was no significant difference (in either direction) for retrieval vs maintenance contrast.
Tables
Name of region Brodmann area number MNI coordinates of the peak/subpeaks t Value of the peak/subpeaks Number of voxels in the cluster Right supplementary motor area BA6 4, 16, 46 15.22 622 −2, 2, 60 12.29 14, 18, 44 12.13 Left inferior frontal gyrus BA44 −50, 6, 0 15.11 201 −58, 6, 16 10.11 Right anterior cingulate BA32 10, 26, 32 14.85 35 Left superior parietal BA40 −48, −28, 48 14.61 1114 −48, −20, 52 12.29 −48, −50, 50 11.97 Right insula/inferior frontal gyrus BA13 30, 24, 2 12.66 281 60, −14, 4 12.07 52, 8, 0 10.16 Left PT/STG −58, −46, 16 12.66 42 Left pons −2, −24, −40 11.49 67 Left anterior insula BA13 −32, 16, 8 10.66 296 −30, 24, 4 10.42 −36, 16, 0 10.29 Left inferior parietal BA7 −24, −66, 46 10.04 77 Left premotor BA6 −30, −4, 62 9.75 20 Left cerebellum −36, −56, −50 9.46 24 Left frontal pole BA11 −24, 44, −14 9.32 15 Left medulla −4, −38, −48 9.23 5 Left frontal BA9 −54, 8, 30 9.11 28 Left posterior temporal lobe −34, −56, 12 9.11 6 Right cerebellum 28, −54, −36 8.90 12 Right posterior STG BA41 46, −36, 16 8.79 12 Right posterior temporal lobe 38, −60, 2 8.79 5 Left superior frontal −24, 4, 66 8.77 8 Right somatosensory/motor area BA1, BA2, BA3 54, −24, 52 8.49 18 Thalamus −10, −16, 0 8.35 6 Left frontal BA9 −50, 4, 40 8.33 5 Right superior parietal BA40 40, −46, 44 8.29 7 The activations (with respect to silent period as baseline) are thresholded at p = 0.05 (corrected for multiple comparisons for the whole brain volume). Only clusters with a size ≥5 voxels are listed. Any subpeaks present in the cluster are also mentioned (peaks are shown in bold letters and subpeaks in nonbold letters).
In this issue
