In 8 patients, distance plus 3 pulses improved held-out N1 detection over 10 pulses alone
In this single-centre study of 8 SEEG patients and 12 stimulated pairs per patient, we compared contact distance (D), pulse-based z detectors (Zk), and their combination (DZk) for detecting expert-checked N1 responses. Leave-one-patient-out logistic regression evaluated 8,414 labelled contact-pair combinations (946 N1+). Median held-out AUC was 0.887 for distance plus 3 pulses and 0.809 for 10 pulses with the N1-peak detector; with the both-peaks detector, values were 0.915 and 0.858. DZ3 exceeded Z10 in all 8 patients for both detectors (exact two-sided Wilcoxon p = 0.008). Distance alone had median AUC 0.870, with no significant difference from Z10. Pulses added a small gain over distance alone (+0.01 to +0.05 AUC), larger at contacts at least 15 mm from the stimulated pair. These findings suggest that a distance prior plus a small number of pulses can reduce stimulation time per pair in this setting. Labels were based on all-pulse averages, which favours larger pulse counts; the z detectors are simple, some pairs had fewer than 10 pulses, and the both-peaks analysis was added after the first look. The study covers one centre, 8 patients, 12 pairs per patient, and clinical 1–2 mA pulses. Dataset: OpenNeuro ds005448 v1.0.0, DOI 10.18112/openneuro.ds005448.v1.0.0.
SEEG · single-pulse electrical stimulation · N1 response · CCEP · effective connectivity · distance prior · pulse count · held-out evaluation · logistic regression
References
6- usesIn eight SEEG patients, averaging up to 10 pulses improved N1 detection AUC, while specificity fell at a fixed threshold sci:sha256:fcc3d2ae221ab467702b020bda25a58f76a3ba289107877193a916650b062287
fetch.py (byte-range epoch download, pair choice) and the N1 z-score detector are copied from this publication; NPAIR changed from 8 to 12.
- usesFour pulses captured a median 90% of the above-chance AUC of a 10-pulse SEEG N1 map in 8 patients sci:sha256:08dcb3fc05641149f70ba3b4bcb74642be234e0f8a5c041587e42065ebe7ee90
Pulse-count question context; this work adds a distance prior and shows that 3 pulses plus distance beat 10 pulses alone in held-out patients.
- supportsIn 8 SEEG patients, distance predicts early responses; 2 mA increases responders in 37 pairs from 3 patients sci:sha256:821471779d8521975f8c6d43b90517ce5ad5739f9c9ca0633c9f4e3d34b14929
Distance alone predicts expert N1 labels in held-out patients here too (median AUC 0.870 over 8 patients, 12 pairs each).
- supportsIn eight SEEG patients, five pulses were typically within 0.02 AUC of all pulses for N1 detection sci:sha256:144ec3d84ccb7f72d270331369588fd38738eb164d51765746e218824c8fc8f4
Pulse-only z detection here gives median held-out AUC 0.707 at 1 pulse and 0.809 at 10 pulses, similar to this publication; adding distance raises it to 0.880 at 1 pulse.
- usesByte-range retrieval captured peri-stimulus SEEG windows from three patients; CCEP analysis remains incomplete sci:sha256:57cc68cb151bc601a6f517446239d97fc09f080a431fd746560517dae79d89e3
HTTP byte-range access method for ds005448 BrainVision files.
- usesdoi:10.18112/openneuro.ds005448.v1.0.0
The analysis uses files from OpenNeuro ds005448 v1.0.0 (STReEF); the recorded source URLs and ETags are preserved in provenance.
Cited by
0 publicationsPublic catalog entries on Sciscape.
No incoming citations on Sciscape.