Article Figures & Data
Figures
- Figure 1.
Sample pulse sequences and experimental paradigm. A, Sample stimuli. Pulse sequences of 7, 9, 11, and 13 were used as experimental stimuli. For each numerosity, there were four different durations (960, 1020, 1080, and 1140 ms), where each duration was subdivided into 60 ms slots. The distribution of pulses to slots was randomized for each stimulus presentation. The first and the last slot of each stimulus always contained a pulse. The stimuli displayed are for illustrative purposes. B, Experimental paradigm. A delayed match-to-numerosity task was used, where two sample stimuli and a mask were presented consecutively. A visual retro-cue that was presented simultaneously with the mask indicated which of the numerosities should be retained for the 12 s delay. After the delay, participants performed a two-alternative forced choice, indicating which of the two test stimuli had the same numerosity as the cued stimulus. The response period was 1.5 s. Please note that the stimulus duration and interstimulus interval changed depending on the stimulus duration, but the onset of each event was locked to coincide with the onset of an image acquisition.
- Figure 2.
A, Mean rate of correct responses across participants (n = 34) for different numerosities in the main WM DMTN task. The figure shows that the WM performance decreases with increasing numerosity. Error bars represent standard derivation (SD). Asterisks indicate statistical significance for pairwise t tests, Bonferroni corrected for multiple comparisons (p < 0.05/6). B, Mean performance across subjects for estimated numerosity in number naming task (mean ± SD). C, True numerosities versus mean numerosity estimations (error bars show SD).
- Figure 3.
A, Brain regions coding information for the memorized estimated numerosities. Group-level results of a t-contrast testing the 12 s WM delay for above-chance prediction accuracy. Brain regions carrying information about memorized scalar magnitudes are as follows: IFG, MFG, PMC, SMA, and SFG. B, Time courses of decoding accuracies of remembered (red) and nonremembered (gray) stimuli for all identified brain regions in the main analysis (Fig. 3A). Error bars indicate standard error of the mean (SEM). The figure shows that, for all clusters depicted in the main analysis, there is more numerosity-specific WM information for the remembered than for the forgotten numerosity, and the information is present throughout the WM delay period. C, Results of the label permutation tests. Five bars are shown for each brain region, respectively. Each bar displays the mean prediction accuracy estimated from the distance to correct order groups. The shade of the bar color, ranging from black to white, depicts the different distance to correct ordering. Black bars indicate the mean prediction performance of the group with the correct linear order, while white bars represent the mean prediction accuracy derived from the most linearly unordered data. Brain regions tested for label permutation are: IFG, MFG, PMC, SMA, and SFG. Error bars indicate SEM.
Tables
- Table 1:
Anatomic label and MNI coordinates of brain areas depicting memorized numerosity information during WM
Peak MNI coordinates Cluster size Anatomical region x y z z-score Prediction accuracy 4557 Left PMC/MI −50 2 52 4.78 0.082 Left SFG −28 0 60 7.74 0.146 SMA −6 10 74 4.48 0.114 1342 Right SFG 32 50 10 4.17 0.135 Right IFG (pars triangularis) 60 24 2 4.17 0.075 Right MFG 40 50 30 3.69 0.069 All results are reported at pFWE-Cluster < 0.05 with a cluster-defining threshold of p < 0.001. Mean prediction accuracy over the delay period is reported. Areas were, where possible, identified using the SPM anatomy toolbox (Eickhoff et al., 2005).
In this issue
