five

fMRIPrep Derivatives for SpaceTop

收藏
OpenNeuro2025-12-12 更新2026-04-11 收录
下载链接:
https://openneuro.org/datasets/ds007070
下载链接
链接失效反馈
官方服务:
资源简介:
# ds005256 - fMRIPrep derivatives This is a BIDS Derivatives dataset resulting from running fMRIPrep on [ds005256](https://openneuro.org/datasets/ds005256). ## Methods fMRIPrep documentation can be found on <https://fmriprep.readthedocs.io/>. A containerized version of fMRIPrep is stored in code/containers/fmriprep_25_1_4.sif. This container was built using ```shell apptainer build code/containers/fmriprep_25_1_4.sif docker://nipreps/fmriprep:25.1.4 ``` Products were generated with the [Slurm](https://slurm.schedmd.com/overview.html) script [`code/run`](code/run). That script uses [Slurm arrays](https://slurm.schedmd.com/job_array.html) to process participants in parallel (one participant per job). Slurm logs (combined `stdout` and `stderr`) are stored in `logs/$SLURM_ARRAY_TASK_ID.log`. Running all participants would entail setting the `array` argument during the call to `sbatch`. ```shell sbatch --array=0-116 code/run ``` Derivatives were saved in the dataset with `git/git-annex` using [`code/save`](code/save). For example, to save sub-0133, run the following. ```shell # NOTE: the argument (116) is the SLURM_ARRAY_TASK_ID, not the participant_label ./code/save 116 ``` ### Defacing To acheive the highest quality derivatives possible, these derivatives were generated on the raw (that is, not defaced) structural scans (sha: 86007170141050f875acfd6daf802d93bd9a7dac). This processing was committed to branch `main-w-faces`. After all participants were added, a new branch was created, `main-w-faces-defacing`, and the defaced images were committed to there (see [`code/remove_faces`](code/remove_faces)). The main branch contains references to only defaced files. ### Environment ```shell $ apptainer --version apptainer version 1.2.2-2.el8 ``` ### Caveats At the time of writing, most fMRIPrep products are not covered by BIDS, and so most of this dataset is covered by a [`.bidsignore`](.bidsignore).
创建时间:
2025-12-12
5,000+
优质数据集
54 个
任务类型
进入经典数据集
二维码
社区交流群

面向社区/商业的数据集话题

二维码
科研交流群

面向高校/科研机构的开源数据集话题

数据驱动未来

携手共赢发展

商业合作