cellmil.datamodels¶

class cellmil.datamodels.WSI(name: str, patient: str, slide_path: Union[str, Path], patched_slide_path: Optional[Path] = None, embedding_name: Optional[Union[str, Path]] = None, label: Optional[Union[str, int, float, ndarray[Any, Any]]] = None, logger: Optional[Logger] = None)[source]¶

Bases: object

WSI object

Parameters:

name (str) – WSI name
patient (str) – Patient name
slide_path (Union[str, Path]) – Full path to the WSI file.
patched_slide_path (Union[str, Path], optional) – Full path to preprocessed WSI files (patches). Defaults to None.
embedding_name (Union[str, Path], optional) – Defaults to None.
label (Union[str, int, float, np.ndarray], optional) – Label of the WSI. Defaults to None.
logger (logging.logger, optional) – Logger module for logging information. Defaults to None.

name: str¶

patient: str¶

slide_path: Union[str, Path]¶

patched_slide_path: Optional[Path] = None¶

embedding_name: Optional[Union[str, Path]] = None¶

label: Optional[Union[str, int, float, ndarray[Any, Any]]] = None¶

logger: Optional[Logger] = None¶

metadata: dict[str, Any]¶

all_patch_metadata: dict[str, Any]¶

patches_list: list[str]¶

patch_transform: Optional[Callable[[Any], Any]]¶

__post_init__()[source]¶: Post-Processing object

_get_metadata() → None[source]¶: Load metadata yaml file

_get_wsi_patch_metadata() → None[source]¶: Load patch_metadata json file and convert to dict and lists

load_patch_metadata(patch_name: str) → dict[str, Any][source]¶

Return the metadata of a patch with given name (including patch suffix, e.g., wsi_1_1.png)

This function assumes that metadata path is a subpath of the patches dataset path

Parameters:: patch_name (str) – Name of patch
Returns:: metadata
Return type:: dict

set_patch_transform(transform: Callable[[Any], Any]) → None[source]¶

Set the transformation function to process a patch

Parameters:: transform (Callable) – Transformation function

process_patch_image(patch_name: str, transform: Optional[Compose] = None) → Tuple[Tensor, dict[str, Any]][source]¶

Process one patch: Load from disk, apply transformation if needed. ToTensor is applied automatically

Parameters:

patch_name (Path) – Name of patch to load, including patch suffix, e.g., wsi_1_1.png
transform (Callable, optional) – Optional Patch-Transformation

Returns:

torch.Tensor: patch as torch.tensor (:,:,3)
dict: patch metadata as dictionary

Return type:

Tuple[torch.Tensor, dict]

get_number_patches() → int[source]¶

Return the number of patches for this WSI

Returns:: number of patches
Return type:: int

get_patches(transform: Callable[[Any], Any]) → Tuple[Tensor, list[dict[str, Any]]][source]¶

Get all patches for one image

Parameters:

transform (Callable, optional) – Optional Patch-Transformation

Returns:

patched image: Shape of torch.Tensor(num_patches, 3, :, :)
coordinates as list metadata_dictionary

Return type:

Tuple[torch.Tensor, list]

load_embedding() → Tensor[source]¶

Load embedding from subfolder patched_slide_path/embedding/

Raises:: FileNotFoundError – If embedding is not given
Returns:: WSI embedding
Return type:: torch.Tensor

__init__(name: str, patient: str, slide_path: Union[str, Path], patched_slide_path: Optional[Path] = None, embedding_name: Optional[Union[str, Path]] = None, label: Optional[Union[str, int, float, ndarray[Any, Any]]] = None, logger: Optional[Logger] = None) → None¶

Modules

`cellmil.datamodels.datasets`
`cellmil.datamodels.model`
`cellmil.datamodels.transforms`	Feature and label transforms for preprocessing.
`cellmil.datamodels.wsi`