科研 Skill

DBS植入物传感数据分析

neuro-dbs-sensing-data

Loads, time-aligns and analyzes sensing data from DBS and bidirectional implants such as Percept and Summit RC+S. Use with implant LFP exports, packet loss or aDBS biomarkers.

发布者 Codar安全审核通过v0.1.0更新于 2026-11-18

SKILL.md

Implant sensing data

Implant exports are not continuous recordings: data come in packets with their own timestamps, packets go missing, several clocks are involved, and ECG and stimulation leak into the LFP. Rebuild a clean, gap-aware time series before any spectral analysis.

0. Load the tool reference

Read references/percept-and-dbs-tools.md: every Percept export section and its timing fields, the perceive toolbox, Lead-DBS and OSS-DBS. Rebuild Percept streams with scripts/percept_timedomain.py, which keeps lost packets as NaN gaps.

1. Load with an established reader

  • Medtronic Percept: the JSON session export has separate sections for survey, streaming and timeline (chronic) data. Community readers exist (for example the Perceive toolbox in MATLAB and Python ports); check their current version against your firmware.
  • Summit RC+S: several JSON files per session; use the open processing pipeline from the investigators who released it (Sellers et al. 2021 describe the data format and timing problems).
  • Record device model, firmware, sensing configuration (contacts, sampling rate, filters) from the export, not from memory.

2. Reconstruct time

  1. Use packet timestamps and packet sizes to place samples; detect gaps (missing packets) and mark them as NaN rather than concatenating.
  2. Convert device time to wall-clock time using the export's reference timestamps; keep the offset and any drift (see neuro-sync-forensics for aligning to external recordings via shared events such as stimulation onset artifacts).
  3. Report total duration, number and length of gaps, and the fraction of data lost.

3. Clean

  • ECG: cardiac artifacts are common, especially with single-sided sensing configurations. Detect QRS complexes (from the LFP itself or an external ECG) and remove with template subtraction; check the spectrum before and after.
  • Stimulation: identify stimulation frequency and its aliases at the sensing sampling rate; analyse bands away from them; mark ramp periods.
  • Movement and telemetry artifacts: mark and exclude.

4. Analyse gap-aware

Compute spectra with Welch on clean contiguous segments only (never across gaps), average by segment length, and report beta peak frequency and power with the segment count. For chronic timeline data, analyse power over time-of-day and against stimulation and medication logs. For aDBS, report the biomarker definition, thresholds, and how often stimulation changed.

Output

Device and configuration, time reconstruction summary (gaps, loss), cleaning steps with before/after spectra, analysis results per segment count.

示例提示词

帮我解析这份Percept PC的BrainSense数据并做beta波段分析。

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

DBS植入物传感数据分析能做什么?

Loads, time-aligns and analyzes sensing data from DBS and bidirectional implants such as Percept and Summit RC+S. Use with implant LFP exports, packet loss or aDBS biomarkers。

如何安装DBS植入物传感数据分析?

在科搭 Codar 左侧点击「插件」,在「组件」页签搜索「DBS植入物传感数据分析」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。

DBS植入物传感数据分析需要哪些权限?

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。

DBS植入物传感数据分析经过官方审核吗?

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。