Calcium Imaging Analysis Guide (Claude Skill)

Gives Claude the decision logic for a calcium-imaging pipeline — which motion-correction and ROI-extraction method fits your preparation, how to correct neuropil contamination, how to compute dF/F, and when deconvolved spike estimates are trustworthy.

   
Type Claude Skill
Supplier Awesome Cognitive and Neuroscience Skills (community OSS, MIT)
Availability GA — one of ~40 research skills in the collection (last upstream push 2026-07-13)
Pricing Free / OSS (MIT)
Capabilities Read/Write — methodology guidance; Claude writes and runs the analysis code locally
Verified works · 2026-08-03
Security cleared · 2026-08-03 — MIT confirmed, repo transferred to NeuroAIHub org (old path still redirects), no advisories

How to install

  • Claude Code — plugin marketplace (installs all skills in the collection): 
    /plugin marketplace add HaoxuanLiTHUAI/awesome_cognitive_and_neuroscience_skills
/plugin install awesome-cognitive-and-neuroscience-skills@awesome-cognitive-and-neuroscience-skills

    Restart Claude Code afterwards. The skills are description-activated — there is no slash command; ask a calcium-imaging question and Claude loads the skill.

  • Claude Code — single-skill alternative. This skill declares a required dependency on the collection’s research-literacy skill, so copy both: 
    git clone https://github.com/HaoxuanLiTHUAI/awesome_cognitive_and_neuroscience_skills
cp -r awesome_cognitive_and_neuroscience_skills/skills/calcium-imaging-analysis-guide ~/.claude/skills/
cp -r awesome_cognitive_and_neuroscience_skills/skills/research-literacy ~/.claude/skills/

    (Project-scoped alternative: copy into .claude/skills/ instead.)

  • Underlying software — the skill is method guidance, not a runner: install whichever extraction package you decide on. Suite2p and Cellpose are pip-installable (pip install suite2p cellpose); CaImAn is distributed for conda/mamba (mamba install -c conda-forge caiman) and MIN1PIPE is MATLAB.

What it does

Steps through the pipeline with a stated rule at each stage:

  • Motion correction — rigid for anesthetized preparations, non-rigid for awake behaving animals; maximum shift capped at ~10% of the field of view.
  • ROI detection — CNMF / sparse NMF for two-photon, CNMF-E for one-photon and miniscope data (where the out-of-focus background is the dominant problem), or Cellpose-based segmentation.
  • Neuropil correction — subtract r × F_neuropil with a coefficient of r ≈ 0.7, before dF/F rather than after.
  • dF/F computation — rolling-percentile baseline, 8th–20th percentile over 30–60-second windows.
  • Spike inference — OASIS, FOOPSI, CASCADE, or MLSpike, with the caveat that deconvolved output is an estimate of firing rate, not a spike train.
  • Quality control — SNR > 3, skewness > 0.5, spatial compactness, and CNN component scoring before a ROI enters the analysis.

Primary use cases: two-photon population imaging, one-photon miniscope recordings in freely moving animals, comparing extraction pipelines on the same dataset.

Notes

AI-generated content — verify before use. The collection’s README states that all skills “were generated by AI from academic papers and other publicly available sources” and that “the content has not been individually verified by human domain experts.” This skill’s front-matter carries review_status: ai-generated. It cites Chen et al. 2013, Pachitariu et al. 2017, Giovannucci et al. 2019, Friedrich et al. 2017 and Zhang et al. 2023 — check the numeric thresholds against those papers and against your own preparation before adopting them.

The named packages carry their own licenses (Suite2p GPL-3.0, CaImAn GPL-2.0, Cellpose BSD-3-Clause); the MIT grant covers the skill text only.

Sources

Installed this tool?

Share feedback — install path, OS, errors, workarounds. The form opens with this tool pre-selected and a link back to this page.