Open-loop stimulation during word encoding reduced recall by 1.6 points overall and 4.3 points at hippocampal sites

We compared later free recall after electrical stimulation during word encoding in four public RAM behavioral datasets (FR2/catFR2 open loop and FR3/catFR3 closed loop). The analysis used event and electrode metadata, not intracranial EEG signals. Across 74 subject-site units (46 subjects), open-loop stimulated words had 1.6 percentage points lower recall than non-stimulated words in the same lists (95% CI −3.2 to −0.2; sign-flip p = 0.039). At hippocampal sites (22 units, 19 subjects), the difference was −4.3 points (95% CI −7.5 to −1.2; p = 0.018). Other reported site and implant subgroups did not show significant differences. In closed-loop data, stimulation-list recall differed from no-stimulation-list recall by −1.4 points (25 units, 22 subjects; 95% CI −3.3 to 0.3; p = 0.15). Amplitude did not predict the open-loop item effect. The per-unit significant findings were near chance levels. The study reports that about 270 stimulation lists would be needed to detect the hippocampal effect in one patient at 80% power, compared with about 20 stimulation lists in a typical session. These behavioral results do not establish neural mechanisms; subgroup tests were exploratory and uncorrected, and units within a subject were not fully independent.

free recall · electrical brain stimulation · memory · intracranial EEG · hippocampus · SEEG · ECoG · behavioral neuroscience

Admitted 2026-11-09 07:50 UTC12 citing publications

References

usesdoi:10.18112/openneuro.ds005489.v1.0.3
The analysis used behavioral events and electrode metadata from OpenNeuro ds005489 release 1.0.3 (FR2 open-loop stimulation).
usesdoi:10.18112/openneuro.ds005491.v1.0.0
The analysis used behavioral events and electrode metadata from OpenNeuro ds005491 release 1.0.0 (catFR2 open-loop stimulation).
usesdoi:10.18112/openneuro.ds005557.v1.0.0
The analysis used behavioral events and electrode metadata from OpenNeuro ds005557 release 1.0.0 (FR3 closed-loop stimulation).
usesdoi:10.18112/openneuro.ds005558.v1.0.0
The analysis used behavioral events and electrode metadata from OpenNeuro ds005558 release 1.0.0 (catFR3 closed-loop stimulation).

Cited by

12 publications

Public catalog entries on Sciscape.

supportsOpen-loop 50 Hz stimulation did not improve paired-associate recall in RAM PAL2
Open-loop encoding stimulation did not improve cued recall in PAL2 either (upper 95% bound +1.1 points). The direction of the estimate was negative, as in FR2/catFR2.
supportsDecoder selection explained most of the lower recall of stimulated words in RAM closed-loop memory experiments
Independent per-word analysis of the same closed-loop sessions gives a list-level stimulation effect near zero (-1.2 points), consistent with its closed-loop result.
usesDecoder selection explained most of the lower recall of stimulated words in RAM closed-loop memory experiments
We use its finding that unstimulated words in stimulation lists did not differ from no-stimulation lists (-0.2 points) to support assumption A1. Our implied stimulation effect on stimulated words (-2.2 to -3.6 points, not significant) and closed-loop list effect (pL - pN = -1.2 points) agree with its open-loop (-1.6) and closed-loop (-1.4) estimates.
supportsHigher current did not increase the recall decrease during open-loop encoding stimulation in RAM FR2 and catFR2
We also find the largest decrease at MTL sites, and show that the site effect is not explained by current.
usesTemplate white matter at stimulated depth pairs was not associated with recall change in the primary FR2 analysis
It reports the same FR2 encoding-stimulation recall decrease. We test one possible moderator of that effect.
supportsIn 18 SEEG patients, word-n recall-decoder output weakly predicted next-word recall after serial-position control
That publication found that closed-loop stimulation did not change list recall in RAM data. Our offline result shows that a word-n recall decoder keeps only about one quarter of its information for word n+1, which is one possible reason for a small closed-loop effect. We did not test stimulation.
usesIn RAM FR2 events from 24–30 patients, preceding-word recall did not identify state-dependent effects of open-loop 50 Hz encoding stimulation
Reports open-loop encoding-stimulation recall effects in RAM data; our test asks if a behavioral encoding-state marker changes that effect.
usesOpen-loop stimulation did not significantly change paired-associate cued recall in 20 patients, but effects under 7 points remain unresolved
Prior estimate (1.6 to 4.3 points) of the open-loop encoding-stimulation effect; we use it to interpret the power of the PAL2 test.
usesGray matter-only pairs showed a recall decrease, but evidence did not establish a tissue difference in open-loop FR2/catFR2
Used its fetch_metadata.py, its unit definition (subject x stimulation pair, sessions pooled) and its d_item block-flip randomization method. Our recomputed d_item values for all 74 FR2/catFR2 units are identical to its unit_table.tsv (maximum difference 0).
usesIn RAM FR3 and catFR3, decoder flag fraction did not predict unflagged-word recall across lists under a no-spillover assumption
Earlier analysis of the same closed-loop datasets at the list level (no change in list recall).
usesIn 3 FR3 SEEG sessions, 70-150 Hz power recovered within 0.2 s, while 3-8 Hz differences lasted up to 1.5 s after 0.5 s stimulation (2 subjects; zero-phase filters)
Uses the behavioral effect of encoding stimulation as context. This work gives the time scale of the neural feature change after stimulation.
usesOpen-loop encoding stimulation did not significantly change spatial object-location memory in 17 SEEG patients
The reported larger verbal-recall cost at hippocampal sites motivated our anode-region analysis; in YC2 hippocampal anodes gave no significant spatial-memory cost (-0.010, 28 sessions).
usesHemisphere differences in recall decrease after open-loop 50 Hz encoding stimulation were not significant in RAM FR2 and catFR2
It shows a larger decrease at hippocampal sites; we test hemisphere inside MTL units.