科研 Skill

相位锁定闭环刺激评估

neuro-phase-locked-stim-eval

Designs and evaluates brain-state- or phase-dependent closed-loop stimulation such as phase-locked TMS or tACS. Use when estimating oscillation phase in real time or checking trigger accuracy.

发布者 Codar安全审核通过v0.1.0更新于 2026-11-18

SKILL.md

Phase-locked stimulation: estimate, trigger, verify

Real-time phase estimates are noisy because the filter and Hilbert transform need future samples. Every system must be evaluated against the phase computed offline with full data, at the moments it actually fired.

1. Choose an estimator

MethodIdeaNotes
AR forward predictionband-pass a past window, fit an autoregressive model, predict ahead, take phase at "now"used for phase-triggered TMS (Zrenner et al. 2018)
Educated temporal predictionlearn the typical cycle length per person, extrapolate from the last peaksimple, fast (Shirinpour et al. 2020)
Endpoint-corrected Hilbert (ecHT)causal Hilbert with end-point correctionlow latency (Schreglmann et al. 2021)
Threshold / peak detectiontrigger on slow-wave down-states or burst amplitudesleep and beta-burst paradigms

Fix the target band per participant from a resting baseline (individual peak frequency) and include a minimum amplitude gate: phase is undefined when the oscillation is weak.

2. Evaluate against offline ground truth

  1. Record the raw signal and every trigger time.
  2. Offline, compute the "true" phase with a zero-phase band-pass and Hilbert on the full recording.
  3. Stimulation artifacts corrupt the data right after a pulse, so take ground truth from sham or catch triggers (estimator fires, no pulse), or from data before the artifact with care.
  4. Report the circular mean phase error, the resultant vector length (consistency), the fraction of triggers within +/-30 or +/-45 degrees of target, and how accuracy depends on amplitude (SNR).

3. Timing chain

Add the hardware delay from decision to pulse (stimulator trigger latency) and compensate it in the prediction horizon; measure it with a loopback. A 10 Hz rhythm moves 3.6 degrees per millisecond.

4. Report

Target band and phase, estimator and parameters, amplitude gate, number of real and sham triggers, phase error mean and spread, consistency, accuracy vs amplitude, and the measured hardware delay.

示例提示词

帮我设计一个在慢波相位触发的闭环听觉刺激实验并评估效果。

所属插件

神经科学面向神经科学研究:EEG/MEG、fMRI、BIDS、spike sorting、连接组、脑机接口解码评估、神经刺激安全与公开神经数据检索。

安装该插件会一并获得同领域的 45 个 Skill / MCP。

同插件的其他科研 Skill

常见问题

相位锁定闭环刺激评估能做什么?

Designs and evaluates brain-state- or phase-dependent closed-loop stimulation such as phase-locked TMS or tACS. Use when estimating oscillation phase in real time or checking trigger accuracy。

如何安装相位锁定闭环刺激评估?

在科搭 Codar 左侧点击「插件」,在「组件」页签搜索「相位锁定闭环刺激评估」并安装;在 Claude Code 或 Codex 中可以发送本页的安装提示词,或下载 Skill 包后解压到 ~/.claude/skills 或 ~/.codex/skills。

相位锁定闭环刺激评估需要哪些权限?

该科研 Skill没有声明额外权限;执行具体操作时仍由 Agent 按自身规则请求确认。

相位锁定闭环刺激评估经过官方审核吗?

是。SpeedSkills 收录的每个版本都会记录内容摘要与权限声明,当前版本已通过安全审核。