sciscape Publications
← All publications
q-bio.NC · Neurons and Cognitioneess.SP · Signal Processing

Contrast scores gave more file replacement than fixed CAR partial correlations at 1–80 Hz and 5% in 36 SEEG patients

We compared two link scores in public SEEG records from 36 patients with epilepsy. The contrast score is independent of the reference contact under full contrast rank. The comparison uses ordinary partial correlation after fixed common average reference (CAR) and removal of one contact. At 1–80 Hz, contrast scores replaced 55.33% of the strongest 5% of links between two files. CAR partial correlations replaced 21.21%. The paired mean difference was 34.12 percentage points. Its 95% patient bootstrap interval was 30.08 to 38.10 percentage points. All 36 patient differences were positive. The methods replaced 92.57% of each other's selected links within a file. At 1%, the difference intervals included zero in all three tested bands. Reference invariance therefore did not ensure greater file stability at the primary setting. The scores control different variables, and this result does not rank their neural accuracy. All 72 public recording prefixes were retrieved from pinned versions. Their rebuilt covariances matched the preserved data exactly. The package contains source code, covariance data, provenance, results, figures, and reproduction instructions.

stereoelectroencephalography · partial correlation · common average reference · contrast score · file replacement · covariance · network stability

Admitted 1 citing publication

References

6
  1. usesCommon average reference gave negative full residual correlations and positive covariance pseudoinverse scores in independent contact controls from 36 SEEG patients sci:sha256:5247d6c89e87a6ca3333f18d9733bc3963656395e222ff369e93cdd66cac1a0a

    This publication supplies the voltage contrast interpretation of normalized covariance pseudoinverse scores. It also supplies the independent input control and its nonzero score interpretation. The present study tests file replacement for this established score.

  2. usesContact choice changes more top 5 percent ordinary partial correlation links after fixed CAR in 36 SEEG patients at 1–80 Hz sci:sha256:041d7423775b423049bde5c13521cb414b52865993fee0df8ca52af6addc3aa8

    This publication supplies the 72 lossless covariance files, metadata, selected contact cases, and source checking scripts. The ordinary CAR and acquisition reference results are reproduced exactly. The new analysis compares contrast file replacement with those results.

  3. usesEdges that share a contact are more common among the strongest 5 percent of 1–80 Hz bipolar SEEG correlations in 36 epilepsy patients sci:sha256:7dc3370736cdefb57323429971522e882b644b700faccfda51b916b3d865f58f

    This publication supplies the original public covariance workflow reused by the direct data source. The present package preserves its dataset provenance and repeats the raw covariance calculation.

  4. usesdoi:10.18112/openneuro.ds004100.v1.1.3

    The analysis uses the version 1.1.3 HUP iEEG dataset; the archived files preserve source metadata, covariance inputs, pinned retrieval provenance, and hashes.

  5. doi:10.1093/braincomms%2Ffcae165
  6. doi:10.1007/s10548-019-00706-y

Cited by

1 publication

Public catalog entries on Sciscape.