Selection by the weaker of two same-sign 30-second correlations reduced mean link retention in 36 SEEG patients at 1–80 Hz with equal 10 percent lists
We tested whether selection by two 30-second correlations improved strong-link agreement in a separate SEEG file. The rule used the weaker absolute half correlation when both signs agreed. The comparison used the full 60-second absolute Pearson correlation. Both rules selected equal 10 percent lists. The primary test used 72 public recordings from 36 patients in OpenNeuro ds004100, version 1.1.3, at 1–80 Hz. Mean retention fell from 61.19 to 58.71 percent. The paired mean change was -2.48 percentage points. The 95 percent patient bootstrap interval was -4.09 to -1.33 points under independent patient sampling. Retention was lower in 30 patients, higher in 3, and equal in 3. All seven planned secondary settings also had a negative mean change. The package preserves code, signed maps, selected links, source pins, metadata, results, and repeat instructions. All 72 source ranges and 221 metadata files were checked. This is a test of recording agreement in the inherited cohort. It does not show neural connection accuracy or clinical benefit.
stereoelectroencephalography · intracranial EEG · Pearson correlation · functional connectivity · link selection · repeatability · epilepsy · patient bootstrap
References
4- usesSpearman gave higher agreement than Pearson between 1–80 Hz bipolar SEEG maps in 18 patients and lower agreement in 18 patients sci:sha256:1c907a005b6735d29f467e29b03c8c661adda48f17e3a259c71b4ab4f94f1e6d
We used the 72 signed Pearson map arrays, the fixed patient cohort, contact selection, and public source records. We copied the Pearson arrays without change. We tested a new link selection rule and agreement endpoint on this same cohort.
- usesTwo bases of the same SEEG contrast space gave different 1–80 Hz correlation participation ratios in 36 patients sci:sha256:901bade82e1972a491fb8b27d87ddf5ae157f439c20fb612d6d24768419473f2
We used the inherited EDF retrieval and decoding code, adjacent bipolar contact rules, source byte range records, and data provenance. These files supported the full public-source rebuild and the independent raw fixture check.
- usesdoi:10.18112/openneuro.ds004100.v1.1.3
The analysis used the HUP iEEG Epilepsy Dataset, OpenNeuro ds004100, version 1.1.3.
- doi:10.1111/epi.16580
Cited by
1 publicationPublic catalog entries on Sciscape.
- usesShaft-pair shrinkage had 1.87 percent higher mean squared z error than global shrinkage in 18 SEEG test patients at 1–80 Hz
We used its 72 signed Pearson map files, patient and contact rules, public source records, rebuild code, and independent raw fixture check. We tested source-only correlation shrinkage on a fixed patient split. This is a new analysis of the same cohort.