neuroscience

---

name: neuroscience description: Supports neuroscience research including brain imaging analysis (fMRI, EEG), neural circuit modeling, cognitive experiment design, and neurological disorder investigation; trigger when users discuss brain regions, neural signals, cognitive tasks, or neuroimaging data.

When to Trigger

Activate this skill when the user mentions:

• fMRI, EEG, MEG, PET, MRI brain imaging
• Neural circuits, synaptic transmission, neurotransmitters
• Cognitive experiments, reaction time, psychophysics
• Brain regions, Brodmann areas, connectome
• Neurological disorders (Alzheimer's, Parkinson's, epilepsy)
• Computational neuroscience, spiking neural networks, Hodgkin-Huxley
• Brain-computer interfaces (BCI), neural decoding

Step-by-Step Methodology

1. Define the neuroscience question - Specify level of analysis (molecular, cellular, circuit, systems, cognitive, behavioral). Identify target brain regions or networks.
2. Experimental design - For imaging studies: specify modality (fMRI for spatial resolution, EEG for temporal resolution, PET for neurochemistry). Design task paradigm with proper controls, counterbalancing, and trial timing (ISI, ITI).
3. Data acquisition guidance - Recommend acquisition parameters: fMRI (TR, voxel size, field strength), EEG (sampling rate, electrode montage, impedance thresholds). Specify preprocessing steps.
4. Preprocessing - fMRI: slice timing, motion correction, normalization (MNI/Talairach), smoothing. EEG: filtering (bandpass), artifact rejection (ICA for eye blinks/muscle), re-referencing. Always report each step and parameters.
5. Analysis - fMRI: GLM for activation, seed-based or ICA for connectivity, MVPA for decoding. EEG: ERP analysis, time-frequency decomposition, source localization. Computational models: implement and fit biophysical or phenomenological models.
6. Statistical inference - Apply appropriate correction for multiple comparisons: cluster-level FWE for fMRI, permutation-based corrections for EEG. Report effect sizes. Use Bayesian approaches when frequentist results are ambiguous.
7. Interpretation - Map results to known neuroanatomy (use atlases: AAL, Desikan-Killiany, Schaefer). Discuss findings in context of established theoretical frameworks. Avoid reverse inference pitfalls.

Key Databases and Tools

• NeuroSynth / Neuroquery - Meta-analytic functional maps
• Allen Brain Atlas - Gene expression and connectivity
• OpenNeuro - Open neuroimaging datasets
• BrainMap - Functional neuroimaging database
• SPM / FSL / AFNI / FreeSurfer - Neuroimaging analysis software
• MNE-Python / EEGLAB - EEG/MEG analysis tools
• NEURON / Brian2 - Neural simulation environments

Output Format

• Brain activation maps with MNI coordinates (x, y, z), cluster size, peak t/z-value.
• ERP waveforms with component labels (N1, P3, N400), latency, and amplitude.
• Time-frequency plots with frequency bands labeled (delta, theta, alpha, beta, gamma).
• Computational model parameters with biological interpretation.

Quality Checklist

• Brain coordinates in standard space (MNI or Talairach) with atlas labels
• Multiple comparison correction method specified and justified
• Sample size adequate for imaging modality (power analysis cited)
• Preprocessing pipeline fully documented (software version, parameters)
• Task design includes appropriate controls and counterbalancing
• Effect sizes reported alongside statistical significance
• Reverse inference explicitly avoided or qualified
• Raw data sharing or availability discussed (OpenNeuro, BIDS format)