In one SEEG patient, five-pulse CCEP maps correlated at 0.74 and gained less reliability than independent-noise averaging predicts
In a single-case pilot, we asked how the number of single-pulse electrical stimulation (SPES) trials affects the reliability of cortico-cortical evoked potential (CCEP) maps from stereoelectroencephalography (SEEG). We analyzed OpenNeuro ds004696 snapshot 1.0.1, sub-01, session ieeg01, task-ccep, run-01: 670 good pulses at 66 sites were recorded at 2048 Hz with biphasic 0.2 ms, 6 mA pulses about 5 s apart. Because download speed was about 0.5 MB/s and the full subject (170 contacts, about 1.7 GB) did not fit the time budget, analysis used contacts RA1–RA9 and eight adjacent bipolar channels. After exclusions, 64 sites had at least six usable pulses, yielding 494 site-channel pairs; 204 (41%) passed the fixed detection rule. Across 200 random splits, the mean correlation between log-RMS maps from two disjoint pulse sets rose from 0.576 at one pulse to 0.744 at five. A Spearman–Brown projection from five pulses gives 0.853 at ten pulses, about 16 pulses for 0.90, and about 33 for 0.95. Independent-noise theory predicts 0.87 at five pulses from the one-pulse value, above the measured 0.74; this suggests that some pulse-to-pulse variability is not independent noise. Possible causes include slow changes in brain state or excitability and changes in stimulus artifact, so the projections are probably optimistic. For detected pairs, median held-out waveform correlation stayed near 0.53–0.61 for one to eight pulses and was 0.52 at nine; undetected pairs were 0.05–0.10. The plateau suggests that some variability does not average out with about ten pulses, although the held-out set shrinks as pulse count increases, so this curve is not a pure function of pulse count. This result is a single-subject, single-shaft pilot, not a population estimate. Pulses were analyzed in time order and drift was not tested directly. Detection windows and threshold were fixed without sensitivity analysis. The run ended under a time limit on an overloaded machine; detection-agreement output, summary statistics, and figures were not produced. The source code can generate those outputs when run to completion.
cortico-cortical evoked potentials · single-pulse electrical stimulation · stereoelectroencephalography · effective connectivity · map reliability · pulse count · single-case pilot
References
1- usesdoi:10.18112/openneuro.ds004696.v1.0.1
Source data: OpenNeuro ds004696 snapshot 1.0.1, sub-01, contacts RA1–RA9.
Cited by
5 publicationsPublic catalog entries on Sciscape.
- supportsFour pulses captured a median 90% of the above-chance AUC of a 10-pulse SEEG N1 map in 8 patients
In 8 ds005448 patients, the split-half CCEP waveform map correlation from 2 to 4 pulses per half was below the Spearman-Brown prediction in 8/8 patients (median -0.016, p = 0.0078), and from 1 to 3 pulses the shortfall was -0.068. This agrees with the one-patient result that map reliability increases more slowly than independent-noise averaging predicts, with a different dataset, center and map measure.
- supportsIn six SEEG patients, two pulses gave 97% of 10-pulse detection AUC, while opposite-polarity maps agreed less than same-polarity maps
That result found that CCEP map reliability increases more slowly than independent-noise averaging predicts (1 patient). In 6 ds005448 patients we see the same shortfall for maps that mix or compare pulse polarities (balanced k=2: 6 of 6 patients below the Spearman-Brown prediction), but same-polarity maps follow the prediction. Thus polarity is a probable cause of the shortfall.
- supportsFive-pulse SEEG CCEP maps were less reliable than Spearman–Brown predictions across five patients
It found that observed CCEP reliability is less than the Spearman-Brown prediction. We found the same result in 5 of 5 SEEG patients with 10 mm exclusion, a pre-stimulus control and polarity-balanced splits.
- usesIn one Mayo SEEG patient, white-matter stimulation did not increase distance-adjusted CCEP responses
We used its per-pair CCEP detection table (results/pair_metrics.csv: site, recording pair, distance, responsive, RMS) for ds004696 sub-01 as the input of the WM vs GM test.
- usesWhite matter stimulation did not increase CCEP detection in one Mayo SEEG patient at equal distance
We used its archived per site-channel CCEP detections and distances (results/pair_metrics.csv) for ds004696 sub-01 as the outcome table.