A flexible calibration rule reduced 20 to 500 ms SEEG forecast error against concentrated calibration at the same budget in eight recordings
A flexible two-to-five pulse calibration rule reduced 20 to 500 ms SEEG forecast error by 3.86% against a concentrated two-or-five rule at the same budget in eight recordings. The 95% person bootstrap interval was 1.92% to 6.10%. The gain against a balanced three-or-four rule was 0.55%, with an interval from -0.79% to 2.29%. The flexible and balanced allocations were identical in five people. Noise-scaled forecasts kept this pattern. An exact descriptive decomposition assigned 95.2% of the mean normalized loss difference against concentration to the pulse-count distribution. The study used 85 sites and 879 bipolar readouts from public OpenNeuro ds004696 version 1.0.1. Rules used two scouts and scored the separate final five pulses. The package preserves code, input data, all loss curves, four figures, source recipes, hashes, and numerical checks. This is exploratory offline reuse. It does not establish delivered-pulse savings or performance of a new stimulation schedule.
SEEG · brain-computer interfaces · stimulation response · calibration · pulse allocation · forecast error · OpenNeuro
References
4- usesAt rate 1, learned weights reduced 20 to 500 ms SEEG forecast error against equal weights but raised error against scaled means in eight recordings sci:sha256:ef0c0959114fb72eaef63b2306de27883d737af4882e1642fc43cc265cd97c7b
We used its all-pulse Gram data, native voltage example, source metadata, pinned source manifest, and retrieval helpers. These inputs enabled exact loss curves for two to five calibration responses. We freshly rebuilt sub-01 and checked the copied input hashes against its archive.
- usesAn offline two-pulse site rank did not show lower 20–500 ms SEEG forecast error at a fixed budget in eight sessions sci:sha256:578806530d9c396756538b53e11352e83c900274801f852ef7211403fa594c3c
We used its two-scout normalized noise priority, fixed person budget, and concentrated two-or-five calibration rule as the comparator. We added flexible marginal-value allocation, a balanced three-or-four comparator, and an exact count-versus-assignment decomposition.
- usesdoi:10.18112/openneuro.ds004696.v1.0.1
Analyzed source dataset, version 1.0.1; the saved dataset metadata records this exact DOI and CC0 license.
- doi:10.1523/JNEUROSCI.2201-22.2023
Cited by
2 publicationsPublic catalog entries on Sciscape.
- usesOffline response noise allocation had an uncertain 20 to 500 ms centered bipolar SEEG forecast gain at three pulses per site in eight recordings
We used its compact all-pulse Gram products, native voltage examples, source provenance, and retrieval helpers. This study changes the question, common scale, comparator, and main budget. It tests response noise against pre-pulse noise at three pulses per site.
- usesA fixed pulse limit increased detection of imposed SEEG gain trends in eight recordings under an exchangeable pulse model
We used the all-pulse extraction and retrieval lineage preserved by the direct source publication. It supplies the compact matrix schema and pinned processed-source receipts.