A paired test did not detect excess beta network change around motor stimulation in one-second SEEG windows from 21 patients
We tested whether beta network changes around motor stimulation exceeded variation between short windows from the same state. We used all 35 recordings from 21 patients in public OpenNeuro SEEG dataset ds006519, version 1.0.0. Each recording supplied two pre-stimulation and two post-stimulation windows. Each window had one second of 13 to 30 Hz signal. The primary rule used signed Pearson edges with no shared contact. We averaged edges, then recordings within each patient. The mean between-state squared change was 0.069670. The mean within-state squared change was 0.070772. Their difference was -0.001102 squared correlation units, with a 95% patient bootstrap interval of -0.002358 to 0.000104. The exact two-sided sign-flip test gave p = 0.1036 under an independent patient model with a symmetric null distribution. Within-state edge-rank agreement was low: mean Spearman correlation 0.0692. The primary test did not detect excess change. This does not prove that stimulation had no effect or that the network was stable. The package preserves event data, network matrices, figures, source hashes, code, and reproduction instructions. Fresh extraction from all 355 verified source files reproduced the event arrays and analysis results exactly. A simulation also shows why the edge median lacks the mean estimator's zero-mean null property.
stereoelectroencephalography · SEEG · beta band · functional connectivity · paired estimator · network change · brain-computer interface · motor stimulation
References
4- usesBipolar SEEG pairs with a shared contact have higher mean absolute beta correlation than the nearest pairs with separate contacts in 21 patients sci:sha256:f9206e6467700c572ec5f73a7622ca0b4d27200ce928e5b1393ab3a0c47567c8
This study uses the verified 35 event files, fixed source manifest, source metadata, retrieval and extraction code, and contact parser. It computes new short-window networks and a paired change endpoint. Fresh source extraction reproduced the inherited event arrays.
- usesdoi:10.18112/openneuro.ds006519.v1.0.0
The analyzed signals and metadata are from OpenNeuro ds006519 version 1.0.0, fixed to the recorded source commit.
- doi:10.1016/j.neuroimage.2026.121822
- supportsdoi:10.1093/braincomms%2Ffcae165
The cited study is the source for the statement that intracranial EEG network estimates depend on re-referencing and connectivity choice.
Cited by
4 publicationsPublic catalog entries on Sciscape.
- usesA baseline rule reduced later beta SEEG correlation error in 1 s windows from 21 patients
We used its frozen 256 Hz event arrays, fixed source identities, and beta window method. Its low rank repeatability motivated the present question. The new test fits a baseline weight and measures later prediction error.
- A fixed 30 mm test of 1 s SEEG windows did not establish greater repeated beta correlation change near motor stimulation in 15 patients
[Legacy relation: reproduces] The native-rate, whole-network mean on the same 35 records was -0.0011013, close to its 256 Hz mean of -0.0011018. This is a processing check on reused signals. It is not independent patient validation.
- usesA fixed 30 mm test of 1 s SEEG windows did not establish greater repeated beta correlation change near motor stimulation in 15 patients
Its repeated-change statistic and result were read before the new spatial scores. The initial plan was amended to test a fixed 30 mm spatial contrast. Its saved record scores provide the comparator for the native-rate check.
- usesSeparate one-second windows did not establish repeatable local beta power changes around motor stimulation in 21 SEEG patients
Its paired network test informed the separation of repeated change from ordinary squared change. This study applies a related product principle to centered log-power changes. It tests a different endpoint. No arrays or code were imported from this publication.
- usesEquivalent bipolar trees gave equal pooled covariance distances within 1e-8 in one-second beta windows from 35 SEEG records with positive definite pooled covariance
We used its between-condition minus within-condition reference design. Our endpoint uses local covariance distances on contact trees and has different channel sets and weights.