voxelops.schemas package

Typed dataclass schemas for procedure inputs, expected outputs, and default configurations. Each procedure defines three dataclasses: Inputs, Outputs, and Defaults.

HeudiConv

HeudiConv schemas: inputs, outputs, and defaults.

class HeudiconvDefaults(heuristic: Path | None = None, bids_validator: bool = True, overwrite: bool = False, converter: str = 'dcm2niix', bids: str | None = 'notop', grouping: str | None = 'all', docker_image: str = 'nipy/heudiconv:1.3.4', post_process: bool = True, post_process_dry_run: bool = False)[source]

Bases: object

Default configuration for HeudiConv.

Parameters:

  • heuristic (Optional[Path], optional) – Path to heuristic.py file (required for conversion), by default None.

  • bids_validator (bool, optional) – Run BIDS validator after conversion, by default True.

  • overwrite (bool, optional) – Overwrite existing output, by default False.

  • converter (str, optional) – DICOM converter to use, by default “dcm2niix”.

  • docker_image (str, optional) – Docker image to use, by default “nipy/heudiconv:1.3.4”.

  • post_process (bool, optional) – Enable post-heudiconv processing, by default True.

  • post_process_dry_run (bool, optional) – Test mode - report only, don’t modify, by default False.

__post_init__()[source]

Ensure heuristic path is Path object if provided.

bids: str | None = 'notop'
bids_validator: bool = True
converter: str = 'dcm2niix'
docker_image: str = 'nipy/heudiconv:1.3.4'
grouping: str | None = 'all'
heuristic: Path | None = None
overwrite: bool = False
post_process: bool = True
post_process_dry_run: bool = False
class HeudiconvInputs(dicom_dir: Path, participant: str, output_dir: Path | None = None, session: str | None = None, heuristic: Path | None = None)[source]

Bases: object

Required inputs for HeudiConv DICOM to BIDS conversion.

Parameters:

  • dicom_dir (Path) – Directory containing DICOM files.

  • participant (str) – Participant label (without ‘sub-’ prefix).

  • output_dir (Optional[Path], optional) – Output BIDS directory, by default None. If None, defaults to dicom_dir/../bids.

  • session (Optional[str], optional) – Session label (without ‘ses-’ prefix), by default None.

  • heuristic (Optional[Path], optional) – Path to heuristic.py file, by default None.

__post_init__()[source]

Ensure paths are Path objects.

dicom_dir: Path
heuristic: Path | None = None
output_dir: Path | None = None
participant: str
session: str | None = None
class HeudiconvOutputs(bids_dir: Path, participant_dir: Path, dataset_description: Path)[source]

Bases: object

Expected outputs from HeudiConv.

Parameters:

  • bids_dir (Path) – Root BIDS directory.

  • participant_dir (Path) – Participant-specific directory (sub-XX/).

  • dataset_description (Path) – dataset_description.json file.

bids_dir: Path
dataset_description: Path
classmethod from_inputs(inputs: HeudiconvInputs, output_dir: Path)[source]

Generate expected output paths from inputs.

Parameters:

  • inputs (HeudiconvInputs) – HeudiconvInputs instance.

  • output_dir (Path) – Resolved output directory.

Returns:

HeudiconvOutputs with expected paths.

Return type:

HeudiconvOutputs

participant_dir: Path

QSIPrep

QSIPrep schemas: inputs, outputs, and defaults.

class QSIPrepDefaults(nprocs: int = 8, mem_mb: int = 16000, output_resolution: float = 1.6, anatomical_template: ~typing.List[str] = <factory>, longitudinal: bool = False, subject_anatomical_reference: str = 'unbiased', skip_bids_validation: bool = False, fs_license: ~pathlib.Path | None = None, docker_image: str = 'pennlinc/qsiprep:latest')[source]

Bases: object

Default configuration for QSIPrep (brain bank standards).

Parameters:

  • nprocs (int, optional) – Number of parallel processes, by default 8.

  • mem_gb (int, optional) – Memory limit in GB, by default 16000.

  • output_resolution (float, optional) – Output resolution in mm, by default 1.6.

  • anatomical_template (List[str], optional) – List of output spaces, by default [“MNI152NLin2009cAsym”].

  • longitudinal (bool, optional) – Enable longitudinal processing, by default False.

  • subject_anatomical_reference (str, optional) – Anatomical reference for longitudinal processing, by default “unbiased”.

  • skip_bids_validation (bool, optional) – Skip BIDS validation, by default False.

  • fs_license (Optional[Path], optional) – Path to FreeSurfer license file, by default None.

  • docker_image (str, optional) – Docker image to use, by default “pennlinc/qsiprep:latest”.

__post_init__()[source]

Ensure fs_license is Path object if provided.

anatomical_template: List[str]
docker_image: str = 'pennlinc/qsiprep:latest'
fs_license: Path | None = None
longitudinal: bool = False
mem_mb: int = 16000
nprocs: int = 8
output_resolution: float = 1.6
skip_bids_validation: bool = False
subject_anatomical_reference: str = 'unbiased'
class QSIPrepInputs(bids_dir: Path, participant: str, output_dir: Path | None = None, work_dir: Path | None = None, bids_filters: Path | None = None)[source]

Bases: object

Required inputs for QSIPrep diffusion preprocessing.

Parameters:

  • bids_dir (Path) – BIDS dataset directory.

  • participant (str) – Participant label (without ‘sub-’ prefix).

  • output_dir (Optional[Path], optional) – Output directory, by default None. If None, defaults to bids_dir/../derivatives/qsiprep.

  • work_dir (Optional[Path], optional) – Working directory, by default None. If None, defaults to output_dir/../work/qsiprep.

  • bids_filters (Optional[Path], optional) – Path to BIDS filters JSON file, by default None.

__post_init__()[source]

Ensure paths are Path objects.

bids_dir: Path
bids_filters: Path | None = None
output_dir: Path | None = None
participant: str
work_dir: Path | None = None
class QSIPrepOutputs(qsiprep_dir: Path, participant_dir: Path, html_report: Path, work_dir: Path, figures_dir: Path)[source]

Bases: object

Expected outputs from QSIPrep.

Parameters:

  • qsiprep_dir (Path) – QSIPrep output directory.

  • participant_dir (Path) – Participant-specific directory.

  • html_report (Path) – HTML report file.

  • work_dir (Path) – Working directory.

  • figures_dir (Path) – QC figures directory.

figures_dir: Path
classmethod from_inputs(inputs: QSIPrepInputs, output_dir: Path, work_dir: Path)[source]

Generate expected output paths from inputs.

Parameters:

  • inputs (QSIPrepInputs) – QSIPrepInputs instance.

  • output_dir (Path) – Resolved output directory.

  • work_dir (Path) – Resolved work directory.

Returns:

QSIPrepOutputs with expected paths.

Return type:

QSIPrepOutputs

html_report: Path
participant_dir: Path
qsiprep_dir: Path
work_dir: Path

QSIRecon

QSIRecon schemas: inputs, outputs, and defaults.

class QSIReconDefaults(nprocs: int = 8, mem_mb: int = 16000, atlases: ~typing.List[str] = <factory>, fs_subjects_dir: ~pathlib.Path | None = None, fs_license: ~pathlib.Path | None = None, docker_image: str = 'pennlinc/qsirecon:latest')[source]

Bases: object

Default configuration for QSIRecon (brain bank standards).

Parameters:

  • nprocs (int, optional) – Number of parallel processes, by default 8.

  • mem_gb (int, optional) – Memory limit in GB, by default 16000.

  • atlases (List[str], optional) – List of atlases for connectivity, by default a long list of atlases.

  • fs_subjects_dir (Optional[Path], optional) – FreeSurfer subjects directory, by default None.

  • fs_license (Optional[Path], optional) – Path to FreeSurfer license file, by default None.

  • docker_image (str, optional) – Docker image to use, by default “pennlinc/qsirecon:latest”.

__post_init__()[source]

Ensure paths are Path objects if provided.

atlases: List[str]
docker_image: str = 'pennlinc/qsirecon:latest'
fs_license: Path | None = None
fs_subjects_dir: Path | None = None
mem_mb: int = 16000
nprocs: int = 8
class QSIReconInputs(qsiprep_dir: Path, participant: str, output_dir: Path | None = None, work_dir: Path | None = None, recon_spec: Path | None = None, datasets: dict[str, Path] | None = None, atlases: List[str] | None = None)[source]

Bases: object

Required inputs for QSIRecon diffusion reconstruction.

Parameters:

  • qsiprep_dir (Path) – QSIPrep output directory.

  • participant (str) – Participant label (without ‘sub-’ prefix).

  • output_dir (Optional[Path], optional) – Output directory, by default None. If None, defaults to qsiprep_dir/../qsirecon.

  • work_dir (Optional[Path], optional) – Working directory, by default None. If None, defaults to output_dir/../work/qsirecon.

  • recon_spec (Optional[Path], optional) – Path to reconstruction spec YAML file, by default None.

  • datasets (Optional[dict[str, Path]], optional) – Dictionary of dataset names and paths, by default None.

  • atlases (Optional[List[str]], optional) – List of atlases for connectivity, by default None.

__post_init__()[source]

Ensure paths are Path objects.

atlases: List[str] | None = None
datasets: dict[str, Path] | None = None
output_dir: Path | None = None
participant: str
qsiprep_dir: Path
recon_spec: Path | None = None
work_dir: Path | None = None
class QSIReconOutputs(qsirecon_dir: Path, participant_dir: Path, html_report: Path, work_dir: Path)[source]

Bases: object

Expected outputs from QSIRecon.

Parameters:

  • qsirecon_dir (Path) – QSIRecon output directory.

  • participant_dir (Path) – Participant-specific directory.

  • html_report (Path) – HTML report file.

  • work_dir (Path) – Working directory.

classmethod from_inputs(inputs: QSIReconInputs, output_dir: Path, work_dir: Path)[source]

Generate expected output paths from inputs.

Parameters:

  • inputs (QSIReconInputs) – QSIReconInputs instance.

  • output_dir (Path) – Resolved output directory.

  • work_dir (Path) – Resolved work directory.

Returns:

QSIReconOutputs with expected paths.

Return type:

QSIReconOutputs

html_report: Path
participant_dir: Path
qsirecon_dir: Path
work_dir: Path

QSIParc

QSIParc schemas: inputs, outputs, and defaults.

class QSIParcDefaults(mask: str | None = 'gm', force: bool = False, background_label: int = 0, resampling_target: str | None = 'data', log_level: str = 'INFO', n_jobs: int | None = 1, n_procs: int | None = 1)[source]

Bases: object

Default configuration for QSIParc (brain bank standards).

Parameters:

  • mask (Optional[str], optional) – Mask to apply during parcellation (“gm”, “wm”, “brain”, or path), by default “gm”.

  • force (bool, optional) – Whether to overwrite existing outputs, by default False.

  • background_label (int, optional) – Label value for background voxels, by default 0.

  • resampling_target (Optional[str], optional) – Resampling strategy (“data”, “labels”, “atlas”, or None), by default “data”.

  • log_level (str, optional) – Logging verbosity (e.g., “INFO”, “DEBUG”), by default “INFO”.

  • n_jobs (Optional[int], optional) – Number of jobs to run in parallel, by default 1.

  • n_procs (Optional[int], optional) – Number of processors to use, by default 1.

background_label: int = 0
force: bool = False
log_level: str = 'INFO'
mask: str | None = 'gm'
n_jobs: int | None = 1
n_procs: int | None = 1
resampling_target: str | None = 'data'
class QSIParcInputs(qsirecon_dir: Path, participant: str, output_dir: Path | None = None, session: str | None = None, atlases: list[parcellate.interfaces.models.AtlasDefinition] | None = None, n_jobs: int | None = None, n_procs: int | None = None)[source]

Bases: object

Required inputs for QSIParc parcellation.

Parameters:

  • qsirecon_dir (Path) – QSIRecon output directory.

  • participant (str) – Participant label (without ‘sub-’ prefix).

  • output_dir (Optional[Path], optional) – Output directory, by default None. If None, defaults to qsirecon_dir parent.

  • session (Optional[str], optional) – Session label (without ‘ses-’ prefix), by default None.

  • atlases (Optional[list[AtlasDefinition]], optional) – List of atlas definitions, by default None.

  • n_jobs (Optional[int], optional) – Number of jobs to run in parallel, by default None.

  • n_procs (Optional[int], optional) – Number of processors to use, by default None.

__post_init__()[source]

Ensure paths are Path objects.

atlases: list[parcellate.interfaces.models.AtlasDefinition] | None = None
n_jobs: int | None = None
n_procs: int | None = None
output_dir: Path | None = None
participant: str
qsirecon_dir: Path
session: str | None = None
class QSIParcOutputs(output_dir: Path)[source]

Bases: object

Expected outputs from QSIParc.

Parameters:

output_dir (Path) – Parcellation output directory.

classmethod from_inputs(inputs: QSIParcInputs, output_dir: Path)[source]

Generate expected output paths from inputs.

Parameters:

  • inputs (QSIParcInputs) – QSIParcInputs instance.

  • output_dir (Path) – Resolved output directory.

Returns:

QSIParcOutputs with expected paths.

Return type:

QSIParcOutputs

output_dir: Path

Module contents

Schemas for procedure inputs, outputs, and defaults.