A contact covariance rule kept SEEG site decisions equal across equivalent bipolar trees in five offline patient tests

We tested whether a fitted SEEG reader kept its site decisions when equivalent bipolar trees used the same contact components. The data came from five patients in OpenNeuro ds004457 version 1.0.2. Four first good pulses per site fitted each reader. Four last good pulses per site supplied 732 tests at 183 sites. The primary window was 20 to 100 ms. A contact incidence matrix set the covariance shrinkage target. A matrix proof showed equal distances for invertible changes between matched trees. The standard diagonal rule changed 117 and 113 decisions for two star layouts. The contact rule changed none. Mean patient accuracy was 40.37% for the contact rule and 40.56% for the diagonal path. All 180 contact conditions kept their decisions. An independent implementation confirmed the result. Full raw extraction and a fresh analysis reproduced the stored inputs and results. The package preserves code, data bins, models, figures, hashes, and rebuild instructions. This is an offline method test in a reused epilepsy cohort. It does not show higher accuracy or live brain-interface benefit.

stereoelectroencephalography · SEEG · bipolar montage · covariance shrinkage · Mahalanobis distance · intracranial stimulation · coordinate invariance · offline classification

Admitted 2026-11-09 19:34 UTC6 citing publications

References

usesA fixed covariance rule gave higher SEEG site accuracy than a diagonal rule in offline tests from five patient sessions sci:sha256:1407cd9ac712a4bcd551a3738cc90c3fab309b871572cbd6fc644b41555f7886
We used its fixed half-shrinkage covariance benchmark and source selection and extraction code. The diagonal rule served as the layout-sensitive comparator.
usesSeven connected difference channels preserved eight-contact SEEG decoder decisions without added error in eight public records sci:sha256:b32011e1fcd7c69767cbcdc0373d03d3ac80e5f6315042a4a4cad70f69293cf2
Its connected-difference transfer argument informed the matched-component scope and graph rejection checks. This study tests regularized site distances on different data.
usesA 90% voltage map rule gave lower 20 to 100 ms site accuracy than calibration accuracy selection in five offline SEEG sessions with one covariance rule sci:sha256:0a0af4bd01d752ba750a0eb10ce0beb69faf4cc0ddcf9c1de1525bacbe542916
We used its contact bins, pinned source metadata, fixed selection rules, and full-reader accuracy values. They defined the data and inherited comparison. The five primary path accuracies were reproduced.
usesChanging the bipolar montage reversed 71 of 296 shaft score orders in four SEEG pulse maps at 20 to 300 ms with fixed contacts sci:sha256:141edb15e4dddc10c4dce91c5b27e74afa9603678bb655c0f30af40de9b1c85d
Its raw-score montage audit informed the coordinate-change question and limits. We extend the test to refitted covariance distances and later site decisions.
usesdoi:10.18112/openneuro.ds004457.v1.0.2
Source intracranial stimulation dataset, exact analyzed version 1.0.2.
usesdoi:10.1523/JNEUROSCI.1325-22.2023
Original study describing the source recordings and cohort.
doi:10.1016/j.neuroimage.2018.08.020
Prior comparison of reference methods for SEEG measurements; it did not test this fitted stimulation-site rule.
doi:10.1093/jrsssb%2Fqkad070
Related work on covariance shrinkage and coordinate transformations; this publication does not implement its estimator.

Cited by

6 publications

Public catalog entries on Sciscape.

Two covariance targets kept SEEG site decisions fixed across full layouts of consistent contact differences but differed on 115 of 536 pulses in four records
[Legacy relation: reproduces] We reproduced all 1,464 early site decisions at shrinkage 0.5 for adjacent and low star layouts. We compared its constant target with a component and time dependent target.
usesA contact covariance rule kept beta distances equal across redundant pair layouts in 35 SEEG records with consistent contact differences
This work supplied the prior contact covariance target for regularization across equivalent SEEG trees. Our distance rule applies that contact geometry to redundant graph outputs and includes a cycle consistency check.
usesCycle checks missed simulated contact errors that changed SEEG site decisions in five offline sessions
The contact incidence matrix and equivalent-layout proof informed the contact fault map. We extend the study to exact decision radii with a fixed original-pair reader.
usesSeparate state sums preserved SEEG site decisions during known invertible reference switches in five offline sessions
This prior study tests equivalent path and star references when each layout is fixed. The present study uses these reference types and tests a switch before a voltage-averaging bin ends.
usesStored band covariance kept fixed state decisions equal across two invertible reference changes on matched contacts in eight offline SEEG tests
Its matched-contact tree test informed the reference-control question and scope. This study instead recovers nonlinear power features from per-window band covariance for a fixed reader.
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
Its test of equal site decisions across equivalent SEEG trees supplied the montage question. We tested a four-window state measure on another public dataset. We do not claim a new general covariance invariance theorem.