WSIMeta¶

tiatoolbox . wsicore . wsimeta . WSIMeta

class WSIMeta(slide_dimensions, axes, level_dimensions=None, objective_power=None, level_count=None, level_downsamples=(1,), vendor=None, mpp=None, file_path=None, raw=None)[source]¶

Whole slide image metadata class.

Parameters:

slide_dimensions (int, int) – Tuple containing the width and height of the WSI. These are for the baseline (full resolution) image if the WSI is a pyramid or multi-resolution.
level_dimensions (list) – A list of dimensions for each level of the pyramid or for each resolution in the WSI.
objective_power (float, optional) – The power of the objective lens used to create the image.
level_count (int | None) – (int, optional): The number of levels or resolutions in the WSI. If not given this is assigned len(level_dimensions). Defaults to None.
level_downsamples (list of float) – List of scale values which describe how many times smaller the current level is compared with the baseline.
vendor (str, optional) – Scanner vendor/manufacturer description.
mpp (float, float, optional) – Microns per pixel.
file_path (Path, optional) – Path to the corresponding WSI file.
raw (dict, optional) – Dictionary of unprocessed metadata extracted from the WSI format. For JPEG-2000 images this contains an xml object under the key “xml”.
axes (str)

slide_dimensions¶

Tuple containing the width and height of the WSI. These are for the baseline (full resolution) image if the WSI is a pyramid or multi-resolution. Required.

Type:: tuple(int)

axes¶

Axes ordering of the image. This is most relevant for OME-TIFF images where the axes ordering can vary. For most images this with be “YXS” i.e. the image is store in the axis order of Y coordinates first, then X coordinates, and colour channels last.

Type:: str

level_dimensions¶

A list of dimensions for each level of the pyramid or for each resolution in the WSI. Defaults to [slide_dimension].

Type:: list

objective_power¶

The magnification power of the objective lens used to scan the image. Not always present or accurate. Defaults to None.

Type:: float

level_count¶: (int): The number of levels or resolutions in the WSI. If not given this is assigned len(level_dimensions). Defaults to len(level_dimensions).

level_downsamples¶

List of scale values which describe how many times smaller the current level is compared with the baseline. Defaults to (1,).

Type:: list of float

vendor¶

Scanner vendor/manufacturer description.

Type:: str

mpp¶

Microns per pixel. Derived from objective power and sensor size. Not always present or accurate. Defaults to None.

Type:: float, float, optional

file_path¶

Path to the corresponding WSI file. Defaults to None.

Type:: Path

raw¶

Dictionary of unprocessed metadata extracted from the WSI format. For JP2 images this contains an xml object under the key “xml”. Defaults to empty dictionary.

Type:: dict

Initialize WSIMeta.

Methods

`as_dict`	Convert WSIMeta to dictionary of Python types.
`level_downsample`	Get the downsample factor for a level.
`relative_level_scales`	Calculate scale of each level in the WSI relative to given resolution.
`validate`	Validate passed values and cast to Python types.

as_dict()[source]¶

Convert WSIMeta to dictionary of Python types.

Returns:: Whole slide image metadata as dictionary.
Return type:: dict
Parameters:: self (WSIMeta)

level_downsample(level)[source]¶

Get the downsample factor for a level.

For non-integer values of level, the downsample factor is linearly interpolated between from the downsample factors of the level below and the level above.

Parameters:

level (float) – Level to get downsample factor for.
self (WSIMeta)

Returns:

Downsample factor for the given level.

Return type:

float

relative_level_scales(resolution, units)[source]¶

Calculate scale of each level in the WSI relative to given resolution.

Find the relative scale of each image pyramid / resolution level of the WSI relative to the given resolution and units.

Values > 1 indicate that the level has a larger scale than the target and < 1 indicates that it is smaller.

Parameters:

resolution (Resolution) – Scale to calculate relative to units.
units (Units) – Units of the scale. Allowed values are: “mpp”, “power”, “level”, “baseline”. Baseline refers to the largest resolution in the WSI (level 0).
self (WSIMeta)

Raises:

ValueError – Missing MPP metadata.
ValueError – Missing objective power metadata.
ValueError – Invalid units.

Returns:

Scale for each level relative to the given scale and units.

Return type:

list

Examples

>>> from tiatoolbox.wsicore.wsireader import WSIReader
>>> wsi = WSIReader.open(input_img="./CMU-1.ndpi")
>>> print(wsi.info.relative_level_scales(0.5, "mpp"))
[array([0.91282519, 0.91012514]), array([1.82565039, 1.82025028]) ...

>>> from tiatoolbox.wsicore.wsireader import WSIReader
>>> wsi = WSIReader.open(input_img="./CMU-1.ndpi")
>>> print(wsi.info.relative_level_scales(0.5, "baseline"))
[0.125, 0.25, 0.5, 1.0, 2.0, 4.0, 8.0, 16.0, 32.0]

validate()[source]¶

Validate passed values and cast to Python types.

Metadata values are often given as strings and must be parsed/cast to the appropriate python type e.g. “3.14” to 3.14 etc.

Returns:: True is validation passed, False otherwise.
Return type:: bool
Parameters:: self (WSIMeta)