sciscape Publications
← All publications
eess.SP · Signal Processing

Median 70–110 Hz power gave higher mean Brier error than block mean power in eight SEEG records with one simulated three-sample contact pulse at ten times early RMS

This study compared mean and median 70–110 Hz power from five 100-sample blocks in eight public SEEG records. Each person had eight fixed bipolar pairs. Early trials supplied the reader fits and the pulse amplitude scale. The later test had 674 trials. The primary test added one three-sample contact pulse at ten times early RMS. Mean Brier error was 0.202070 for block mean power and 0.215147 for block median power. The paired difference was +0.013077. The descriptive 95% person bootstrap range was +0.002438 to +0.023904. The exact sign-flip reference was 0.1015625. Most of the error difference was present on intact inputs. Median power gave smaller output changes from the pulse. It gave lower error at the largest tested amplitude. A rank proof limits median power when only one block changes. The package contains inputs, code, models, predictions, source checks, figures, and repeat instructions. These results apply to reused clinical records and synthetic pulses in an offline test.

stereoelectroencephalography · signal processing · gamma power · median estimator · brain-computer interface · synthetic artifact · Brier error

Admitted 0 citing publications

References

7
  1. usesGaps of 50 samples gave higher later Brier error than gaps of 1 sample with linear repair and training on intact data in eight reused SEEG records at 20 percent loss sci:sha256:3c89bd7257f7bec674ddc63c692c0bc98521a393ec34f50a07a13af56427ad8c

    This study uses the exact digital bipolar buffers, trial and contact selection, source metadata, EDF range reader, and logistic fitting equations from this publication. New code computes gamma-only readers and contact pulse tests.

  2. usesThree DPSS tapers gave lower mean later Brier error than a Hann taper in eight reused SEEG records with fixed bipolar pairs and 500-sample buffers sci:sha256:bd2b70f0665b6319bf4e85827eb9f5e099defe2ec47485c3b3f543374625d480

    This publication supplies the earlier waveform and selection lineage and power-equation context. Its prior taper results are not new observations in the present test.

  3. usesEight offline SEEG readers with median gamma power had 0.500 mean worst contact Brier error under simulated 1 mV, 100 Hz faults of 250 samples sci:sha256:89014e7f81930682f7da691822959c491b7c2ec089eaaced910befc99bec8aaa

    This related gamma-only study supplies context for the error tradeoff and block-power bounds. It used worst-contact error, 1 mV faults, and C=1. The present study adds average-contact error for three-sample pulses scaled by early RMS, with C=0.1.

  4. usesMedian block powers gave higher task-state accuracy than mean powers for simulated 1 mV, 10-sample SEEG faults on one contact in eight recordings sci:sha256:7e4be46c9c145b5de38d258bb39ea502a49c77fca52c1e1d18681de88d768b0a

    This related study supplies context for median block power, the intact-data cost, and conditional rank bounds. It used three bands and ten-sample 1 mV contact pulses. It was read after the present primary calculations and did not change the plan or fits.

  5. usesdoi:10.18112/openneuro.ds004473.v1.0.2

    The study uses SEEG task records from OpenNeuro ds004473 version 1.0.2.

  6. doi:10.1088/1741-2552%2Facae0a
  7. doi:10.1088/1741-2552%2Fad1385

Cited by

0 publications

Public catalog entries on Sciscape.

No incoming citations on Sciscape.