nifreeze.data.pet package

PET data representation

This submodule implements data structures and I/O utilities for PET data.

Data Representation. The PET class requires a dataobj that can be an array-like object. The class instantiation requires midframe and total_duration value data to be provided. Both these values are computed from the frame start time (FrameTimesStart in BIDS terms). The from_nii() function takes a temporal file parameter, which is expected to be a JSON file containing the frame start time. Ultimately, the midframe and total duration values are computed by the compute_temporal_markers() method.

class nifreeze.data.pet.PET(self, dataobj: 'np.ndarray' = None, affine: 'np.ndarray' = None, brainmask: 'np.ndarray | None' = None, motion_affines: 'np.ndarray | None' = None, datahdr: 'nb.Nifti1Header | None' = None, filepath: 'Path' = NOTHING, midframe: 'Any' = None, total_duration: 'Any' = None) → None[source]

Bases: BaseDataset[ndarray]

Data representation structure for PET data.

Method generated by attrs for class PET.

classmethod from_filename(filename: Path | str) → Self[source]

Read an HDF5 file from disk.

lofo_split(index)[source]

Leave-one-frame-out (LOFO) for PET data.

Parameters:

index (int) – Index of the PET frame to be left out in this fold.

Returns:

  • (train_data, train_timings) (tuple) – Training data and corresponding timings, excluding the left-out frame.

  • (test_data, test_timing) (tuple) – Test data (one PET frame) and corresponding timing.

midframe: np.ndarray

A (N,) numpy array specifying the midpoint timing of each sample or frame.

set_transform(index: int, affine: ndarray, order: int = 3) → None[source]

Set an affine, and update data object and gradients.

to_filename(filename: Path | str, compression: str | None = None, compression_opts: Any = None) → None[source]

Write an HDF5 file to disk.

total_duration: float

A float representing the total duration of the dataset.

nifreeze.data.pet.from_nii(filename: Path | str, temporal_file: Path | str, brainmask_file: Path | str | None = None) → PET[source]

Load PET data from NIfTI, creating a PET object with appropriate metadata.

Parameters:

  • filename (os.pathlike) – The NIfTI file.

  • temporal_file (os.pathlike) – A JSON file containing temporal data. It must at least contain FrameTimesStart data.

  • brainmask_file (os.pathlike, optional) – A brainmask NIfTI file. If provided, will be loaded and stored in the returned dataset.

Returns:

A PET object storing the data, metadata, and any optional mask.

Return type:

PET

Raises:

RuntimeError – If FrameTimesStart is not provided (BIDS requires it).

Submodules