In 28-subject RAM FR2 depth-anode data, encoding stimulation reduced recall of stimulated words but showed no detectable change in the next word
We used the events files of OpenNeuro ds005489 (FR2) and ds005491 (catFR2) to measure how far in time the recall effect of encoding stimulation continues. On stimulation lists, 4.6 s trains of 50 Hz stimulation are applied to alternating blocks of 2 words. We compared each word category with words of no-stimulation lists at the same serial position, in each subject. In FR2 with depth (SEEG) anodes (28 subjects, 49 sessions), recall of stimulated words decreased by 4.1 percentage points (95% CI -6.8 to -1.5; sign-flip p = 0.004; Holm p = 0.018). Recall of the next unstimulated word, which starts a median 0.61 s after the stimulation ends, did not change (-0.7 points; 95% CI -3.4 to +2.2; p = 0.66; this was the primary test). The lag-2 word also did not change (-0.8 points; p = 0.62). The lag-1 effect was smaller than the effect on stimulated words (+3.5 points; p = 0.024, post hoc). The catFR2 data (9 subjects) gave the same directions but had too little power for a test on its own. Thus, the behavioral cost of this open-loop protocol stays on the items that are shown during the stimulation train. The CI does not exclude a lag-1 decrease of at most approximately 3 points. The analysis uses events only, numpy only, and a plan that we wrote before we computed the effects. The package contains the code, the file manifest with MD5 values, the subject-level results and the figure.
memory · free recall · encoding stimulation · open-loop stimulation · SEEG · memory BCI
References
3- uses50 Hz open-loop depth stimulation during word encoding decreased later recall in ds005489; surface stimulation showed no effect sci:sha256:4636af3ff0cc862353979c5b008d8dee3803276e9ab70b7bea23c76a9bba0bf0
We use its archived words table for the anode type (depth or surface) of each stimulation list. Our ON-word result agrees with its result that encoding stimulation decreases recall in FR2, and we extend it to the time course after stimulation.
- usesdoi:10.18112/openneuro.ds005489.v1.0.3
Primary FR2 dataset; the analysis used its events files.
- usesdoi:10.18112/openneuro.ds005491.v1.0.0
catFR2 replication dataset; the analysis used its events files.
Cited by
6 publicationsPublic catalog entries on Sciscape.
- supportsOpen-loop 50 Hz encoding stimulation through depth anodes decreased correct recall but did not increase recall intrusions
In the same depth-anode FR2 sessions, correct recall per list was lower on stimulation lists (-0.36 words per list), which agrees with its recall decrease.
- usesOpen-loop 50 Hz encoding stimulation through depth anodes decreased correct recall but did not increase recall intrusions
We used its archived events-file key manifest with MD5 sums (all 80 files matched), its per-list anode-type tables for the depth-anode session inclusion, and its statistics functions (sign-flip, exact Wilcoxon, bootstrap).
- usesIn RAM FR2 depth-anode data, encoding stimulation did not measurably delay recalled words
Prior result on the stimulation recall effect; this work adds output order, latency and first-recall tests.
- supportsIn three SEEG patients, most unstimulated pairs showed no detectable difference from 50 ms after logged stimulation offset
Same catFR2 protocol. At unstimulated pairs, the recording had no detectable difference from 50 ms after the train end. This agrees with a behavioral cost that stays on words shown during the train.
- usesIn RAM FR2 events from 24–30 patients, preceding-word recall did not identify state-dependent effects of open-loop 50 Hz encoding stimulation
Same FR2 open-loop data. That result gives the word-level stimulation effect; we add a test of its dependence on the recall of the preceding word.
- supportsOpen-loop 50 Hz stimulation showed no reliable onset-versus-sustained recall cost difference in RAM FR2
With no-stim lists as the reference, stimulated words lose about 3.6 pp and the word after a train loses about 0 pp. This agrees with a cost only on stimulated words. The post-hoc pre-train word cost (-3.4 pp) adds a possible exception for the word before a train.
- usesIn RAM FR3 and catFR3, decoder flag fraction did not predict unflagged-word recall across lists under a no-spillover assumption
Open-loop evidence that stimulation does not change recall of the next unstimulated word, which supports the no-spillover assumption.