Selection by correlation gave more normalized gamma voltage dimensions than selection by power for four SEEG channels in 33 patients
We compared four bipolar SEEG channels from eight candidate contacts on each of 354 shafts in 33 patients. We used public interictal data from OpenNeuro ds004100 version 1.1.3. We selected pairs with the first 60 seconds of file A. We tested them with the first 60 seconds of file B. Selection by correlation gave 3.94 normalized gamma voltage dimensions. Selection by power gave 3.27. The paired patient mean increase was 0.670. Its 95 percent bootstrap interval was 0.626 to 0.711. The mean power ratio was 0.579. A fixed disjoint-pair rule gave 3.90 dimensions and a power ratio of 0.719. These ratios are geometric means across shafts and patients. The score measures voltage redundancy. It does not establish better task decoding or stimulation. The package preserves code, input covariance, all subset scores, figures, source checks, and replay instructions.
SEEG · bipolar channels · channel selection · voltage covariance · participation ratio · signal redundancy · intracranial EEG
References
4- usesEight high-gamma SEEG pairs selected with ten seconds had 76.7% mean retention across two interictal records in 36 patients sci:sha256:4947f6220434c945bfb5f749fa56c65ce926ce8b7742925c1b030438e32ad73f
The fresh-source verification uses the EDF header parser preserved from this publication in provenance/parent/upstream_extract.py. The original code and source attribution are included.
- usesLinks with a shared contact had a larger share of strong local gamma correlations over 60 seconds in 33 SEEG patients sci:sha256:77fdc837b559f7b809e6d187400a439fa1d2e22314b585079cbb13cc1ef1b002
This study uses all 72 contact covariance inputs, source contact rules, and pinned source provenance from this publication. It extends the strong-link audit to a four-output selection benchmark with test scores in the second file.
- usesdoi:10.18112/openneuro.ds004100.v1.1.3
The analysis used the first 60 seconds of two interictal files per patient from OpenNeuro ds004100 version 1.1.3; the saved source metadata records this exact version and CC0 license.
- doi:10.3389/fnsys.2018.00030
Cited by
2 publicationsPublic catalog entries on Sciscape.
- usesLow correlation selection increased later feature dimensions in eight SEEG records. Fixed readers showed no clear key error decrease with eight pairs and 64 labels.
Its correlation selection result in resting SEEG motivated a task test of dimension scores. We used none of its data. We tested power features, eight pairs and later key error, which differ from its voltage features and four-channel test.
- usesMaps fitted to one SEEG record retained 91.9 percent mean gamma energy in another record after common mean removal in 33 patients
We used its rule for the lowest complete eight-contact block on each shaft. This fixed the contact geometry before map fitting.