Neural Noise- Phase-lagged transcranial alternating current stimulation (tACS) applied between executive and default mode networks modulates working memory.1
- The finding informs non-invasive neuromodulation and network-level cognitive protocols (Nature).1 1
Weekly enrichment (2026-07-20)
- The source study is published in Scientific Reports (Nature Portfolio, 2025) and used simultaneous transcranial alternating current stimulation and functional MRI (tACS-fMRI) to test whether phase-lagged stimulation between the central executive network (CEN) and default mode network (DMN) modulates working memory.2
- Twenty-six healthy volunteers participated (nine female; mean age 24.1 ± 2.9 years), approved by the Korea University IRB; data from six participants were excluded for excessive head motion during fMRI.2
- Participants performed a modified Sternberg task with a 7-item memory load of letter-digit combinations (each item shown 700 ms with a 150 ms inter-stimulus interval, followed by a 9 s retention period), and tACS was delivered only during the retention phase.2
- Stimulation used individually customized cross-frequency coupling — theta/alpha phase coupled to high-gamma amplitude, with frequencies set from each subject’s EEG power-spectrum peaks — delivered via an MR-compatible multi-channel high-definition electrical stimulator.2
- CEN nodes (right/left posterior parietal cortex and right/left dorsolateral prefrontal cortex) and DMN nodes (medial prefrontal cortex and posterior cingulate cortex) were stimulated with a phase lag of either 45° (partially in-phase, hypothesized facilitation) or 180° (anti-phase, hypothesized inhibition).2
- The design ran five consecutive fMRI runs (no-tACS, 45°, no-tACS, 180°, no-tACS) with the 45°/180° order counterbalanced, and participants were split into Fast and Slow groups by median split of reaction time in the initial no-tACS run.2
- In the Fast Group, 180° anti-phase tACS significantly slowed Sternberg reaction times versus no-tACS (758.36 ms vs 691.88 ms; t(9) = −2.98, p = 0.016, FDR-corrected) and versus the 45° condition (690.80 ms; t(9) = −2.93, p = 0.017), demonstrating phase-dependent inhibition.2
- Right-hemisphere hippocampus showed significantly higher activation in the Fast versus Slow group during the 180° condition relative to no-tACS, and the 45° condition enhanced CEN–DMN functional connectivity, indicating non-invasive scalp tACS can selectively engage a task-relevant deep structure.2
- The authors position phase-lagged, network-wise tACS as a non-invasive alternative to two-region protocols and relate it to temporal-interference stimulation for deep targets, though they caution that limited sample size and Fast/Slow subgrouping constrain statistical power.2