Complete AI Training
Sign inGet my AI kit

Your job's AI kit

Get your AI kit

Tell us who you are and what you do. We show you your kit right away and email you the link: skills, prompts, AI agents, MCP servers and courses for your job.

500+ jobs ready, and we make a kit for any other job. No payment needed to look.

Share

AI agent for neuroscientists

Imaging Preprocessing Quality Agent

Have every subject either pass quality checks or have a documented exclusion before group analysis.

Imaging Preprocessing Quality Agent: what goes in, what the agent does and what you get

What it does

Head motion and artifacts in brain imaging can quietly bias results, and are often discovered when a group result looks odd. This agent runs the lab's preprocessing on each new subject and checks the output right away. It reads motion parameters, framewise displacement, signal to noise, registration quality and any spikes. It flags subjects with too much motion or poor registration, and reruns preprocessing with alternative settings such as different smoothing, motion correction or a different registration method. It then rechecks the quality measures, and lists subjects that still fail. It records exclusions with reasons and the numbers behind them. It never excludes anyone alone. The scientist approves them. Edge case: a short run with high motion early is trimmed only if the analysis plan allows it.

How it works

Follow the arrows from top to bottom. The orange dashed arrow is the loop: when a check fails, the agent goes back and tries again.

Start and resultWhat it doesA check on its own workWaits for your OKGoes back and retries
Yes, continueYes, continueApprovedNoNo 1 STARTS WHEN New subject data arrive 2 USES A TOOL Run the standard preprocessing 3 USES A TOOL Compute motion, signal and registration metrics 4 CHECKS THE RESULT Do the metrics pass the plan limits? If not: rerun with alternative smoothing, motioncorrection or registration. Back to step 2. 5 DOES Recompute metrics on the rerun 6 CHECKS THE RESULT Does the rerun pass? If not: mark the subject as a proposed exclusion withthe numbers. Back to step 4. 7 DOES Add the subject to the quality table 8 DOES Summarize exclusions and reasons for the study 9 YOU APPROVE Scientist approves exclusions 10 RESULT Quality table and exclusion log
Read the steps as a list
  1. New subject data arrive
  2. Run the standard preprocessing
  3. Compute motion, signal and registration metrics
  4. Do the metrics pass the plan limits?If not: rerun with alternative smoothing, motion correction or registration. Back to step 2.
  5. Recompute metrics on the rerun
  6. Does the rerun pass?If not: mark the subject as a proposed exclusion with the numbers. Back to step 4.
  7. Add the subject to the quality table
  8. Summarize exclusions and reasons for the study
  9. Scientist approves exclusionsThe agent waits here for your OK.
  10. Quality table and exclusion log

How it decides

A subject passes when motion and registration metrics are within the plan limits. Failing subjects get alternative settings once before exclusion is proposed.

  • Flag a subject with mean framewise displacement above 0.5 mm
  • Flag registration with overlap below the set score
  • Rerun each failing subject once with alternative settings
  • Trim volumes only if the analysis plan permits

Make it yours

Every agent is a starting point. You choose these settings for your own situation.

  • Motion limit (default 0.5 mm)
  • Registration quality limit
  • Alternative settings list
  • Quality metrics reported
  • Trimming rule

What keeps you in control

It always asks you first

  • Every exclusion
  • Any change to the pipeline for the whole study

Hard limits

  • Never delete raw data
  • Never exclude before approval

It stops when

  • Done: all subjects pass or are excluded and approved
  • Stop: raw data incomplete

Set it up

We guide you through the set-up, step by step

Members get the full set-up guide for this agent. No technical skills needed: you copy, paste and upload.

10 minto set it up in your AI
5 AIsChatGPT, Claude, Copilot, Gemini, Grok
  • One set of instructions to paste into your AI, with the clicks for ChatGPT, Claude, Microsoft 365 Copilot, Gemini and Grok
  • The agent then walks you through connecting your own data, one source at a time
  • A downloadable copy with the flow chart, the rules and the full guide
Get access to this agent

An example run

What happensOf 32 subjects, 5 had mean displacement above 0.5 mm. The agent reran them with stricter motion correction. Two dropped to 0.38 and 0.44 mm and passed. Three stayed above 0.6 mm. Subject 19 also had poor registration, which an alternative method fixed. The scientist approved excluding the three subjects with the logged numbers.

More agents for neuroscientists