neurodesign-plus 2.0.0

Trial-aware neurodesign-plus 2.0 for fMRI experimental design optimisation

neurodesign-plus

neurodesign-plus is an extended and maintained fork of the original neurodesign package for fMRI experimental design optimisation.

In addition to the base package workflow, neurodesign-plus adds support for:

1. fixed user-defined event orders via order,
2. variable stimulus durations via stimuli_durations,
3. transition-specific ITIs via conditional_ITI,
4. probabilistic template-based ordering via order_keys, order_probabilities, and order_length.

Current version

Version 2.0 is a breaking release: experiments are now built from trials (each holding one or more events) instead of a flat event list, so a few options were renamed. 1.0.x scripts need these renames:

1.0.x	2.0
stimuli_durations	event_durations
conditional_ITI	event_transition_interval / inter_trial_interval
order_keys / order_probabilities / order_length	trial_templates / trial_template_probabilities / n_conceptual_trials

Experiment, Design, and Optimisation are unchanged. See the migration guide to update old code, or pin neurodesign-plus<2.0 to keep the old behavior.

Installation

Install the published package with:

python -m pip install neurodesign-plus

Documentation

• neurodesign-plus documentation: neurodesign-plus on Read the Docs
• User guide: docs/index.md
• Installation notes: docs/installation.md
• Migration guide (version-1 to version-2 migration details): MIGRATION_2.0.md
• Technical manual: manuals/TECHNICAL_CHANGES.md
• Metrics guide: manuals/METRICS.md
• Tutorials: docs/tutorials.md

Current API Example

from pathlib import Path
import json

from neurodesign import Experiment, Optimisation, report

exp = Experiment(
    TR=2.0,
    n_trials=8,
    P=[0.5, 0.5],
    C=[[1, -1]],
    rho=0.3,
    n_stimuli=2,
    event_durations=1.0,
    trial_start_interval=0.5,
    post_event_interval=0.2,
    inter_trial_interval=2.0,
    resolution=0.1,
    seed=7,
)

optimisation = Optimisation(
    experiment=exp,
    weights=[0.0, 0.5, 0.25, 0.25],
    preruncycles=1,
    cycles=1,
    optimisation="simulation",
    G=2,
    I=1,
    outdes=1,
    convergence=1,
    seed=101,
    folder=Path("output") / "readme_example",
)
optimisation.optimise()
design = optimisation.selected_design(0)

report.make_report(optimisation, Path("report.pdf"))
Path("schedule.json").write_text(
    json.dumps(design.export_payload(), indent=2), encoding="utf-8"
)
Path("specification.json").write_text(
    json.dumps(exp.export_specification(), indent=2),
    encoding="utf-8",
)

Repository Layout

docs/        Sphinx documentation sources
manuals/     Included long-form manuals
neurodesign/ Package source code
tests/       Automated regression and release-audit tests
tutorials/   Executed tutorial notebooks
validation/  Validation runners and reproducibility helpers

Tutorials

The tutorial collection includes:

• three overview tutorials in tutorials/,
• four base-function tutorials in tutorials/base_functions/,
• four feature-focused tutorials in tutorials/new_functions/.

The Read the Docs tutorials page is the best place to browse the current notebook set and their intended learning progression.

Credits

This project is a maintained fork of the original neurodesign package by Joke Durnez and the Neuropower team.

Author contributions (CRediT):

• Atharv A. Umap — Conceptualization, Software, Validation
• Caroline J. Charpentier — Project administration
• Valentin Guigon — Conceptualization, Methodology, Resources, Software, Supervision, Validation, Writing