CloseChoice/neuroimaging-go-brrrr

[Dataset] Add AOMIC-PIOP2 (ds002790) - Amsterdam Multimodal MRI (226 subjects)

Open

#14 创建于 2025年12月9日

在 GitHub 查看
 (0 评论) (0 反应) (0 负责人)Python (3 fork)auto 404
datasetenhancementhelp wanted

仓库指标

Star
 (1 star)
PR 合并指标
 (PR 指标待抓取)

描述

Prerequisites

Complete AOMIC-PIOP1 first - PIOP2 uses the same pattern.

Dataset Info

Field Value
Name AOMIC-PIOP2 (Population Imaging of Psychology 2)
Source OpenNeuro ds002790
Paper Snoek et al., Scientific Data 2021
License CC0 (Public Domain)
Subjects 226
Format BIDS
HuggingFace Target hugging-science/aomic-piop2

Description

AOMIC-PIOP2 is similar to PIOP1 but scanned on updated scanner hardware ("Achieva" vs "Intera"). Contains rich task-based fMRI with different experimental paradigms.

Data includes:

  • T1-weighted structural MRI
  • Diffusion-weighted MRI
  • Resting-state fMRI
  • Task-based fMRI (movie watching, different tasks than PIOP1)
  • Physiological recordings

Why This Matters

  1. Scanner upgrade validation - Test reproducibility across hardware
  2. Different tasks - Complementary to PIOP1
  3. Complete AOMIC - Third piece of the collection

Exact Schema

Same as PIOP1 - reuse get_aomic_piop1_features() pattern.

Files to Create

src/bids_hub/datasets/aomic_piop2.py       # Dataset module (copy PIOP1, adjust)
scripts/download_aomic_piop2.sh            # Download script
tests/test_aomic_piop2.py                  # Tests
docs/dataset-cards/aomic-piop2.md          # Dataset card (follow arc-aphasia-bids.md pattern)

Implementation Notes

  • Nearly identical to PIOP1 implementation
  • Can share validation code
  • num_shards=226

Acceptance Criteria

  • Download script works
  • Validation passes
  • Build succeeds
  • Tests pass
  • Dataset uploaded to hugging-science/aomic-piop2
  • HuggingFace README.md with proper frontmatter, usage examples, and citation
  • docs/dataset-cards/aomic-piop2.md added (follow arc-aphasia-bids.md pattern)

Resources

Citation

Same as PIOP1 - Snoek et al., 2021

贡献者指南