fusion_tools package

• fusion_tools.components package
  • fusion_tools.components.maps module
    • ChannelMixer
      • ChannelMixer.add_color_selector_tab()
      • ChannelMixer.gen_layout()
      • ChannelMixer.get_callbacks()
      • ChannelMixer.load()
      • ChannelMixer.process_frames()
      • ChannelMixer.updat_layout()
      • ChannelMixer.update_channel_mix()
      • ChannelMixer.update_slide()
      • ChannelMixer.update_tab_style()
    • MapComponent
    • MultiFrameSlideMap
      • MultiFrameSlideMap.gen_layout()
      • MultiFrameSlideMap.load()
      • MultiFrameSlideMap.process_frames()
      • MultiFrameSlideMap.update_layout()
    • SlideImageOverlay
      • SlideImageOverlay.get_image_bounds()
      • SlideImageOverlay.to_dict()
    • SlideMap
      • SlideMap.add_manual_roi()
      • SlideMap.download_manual_roi()
      • SlideMap.export_image_overlay_bounds()
      • SlideMap.gen_image_mover()
      • SlideMap.gen_layout()
      • SlideMap.get_callbacks()
      • SlideMap.get_click_popup()
      • SlideMap.get_image_overlay_popup()
      • SlideMap.get_namespace()
      • SlideMap.get_scale_factors()
      • SlideMap.load()
      • SlideMap.make_geojson_layers()
      • SlideMap.move_image_overlay()
      • SlideMap.update_image_overlay_transparency()
      • SlideMap.update_layout()
      • SlideMap.update_slide()
      • SlideMap.update_vis_session()
      • SlideMap.upload_shape()
  • fusion_tools.components.tools module
    • GlobalPropertyPlotter
      • GlobalPropertyPlotter.extract_all_properties()
      • GlobalPropertyPlotter.extract_property_info()
      • GlobalPropertyPlotter.gen_layout()
      • GlobalPropertyPlotter.gen_scatter_plot()
      • GlobalPropertyPlotter.gen_umap_cols()
      • GlobalPropertyPlotter.gen_violin_plot()
      • GlobalPropertyPlotter.generate_plot()
      • GlobalPropertyPlotter.generate_property_dict()
      • GlobalPropertyPlotter.get_callbacks()
      • GlobalPropertyPlotter.load()
      • GlobalPropertyPlotter.parse_filter_divs()
      • GlobalPropertyPlotter.select_data_from_plot()
      • GlobalPropertyPlotter.update_drop_type()
      • GlobalPropertyPlotter.update_filter_properties()
      • GlobalPropertyPlotter.update_properties_and_filters()
      • GlobalPropertyPlotter.update_property_selector()
    • HRAViewer
      • HRAViewer.gen_layout()
      • HRAViewer.get_callbacks()
      • HRAViewer.get_organ_table()
      • HRAViewer.load()
      • HRAViewer.update_hra_viewer()
    • MultiTool
    • OverlayOptions
      • OverlayOptions.add_filter()
      • OverlayOptions.add_structure_line_color()
      • OverlayOptions.adv_update_overlays()
      • OverlayOptions.create_layer_from_filter()
      • OverlayOptions.export_layers()
      • OverlayOptions.gen_layout()
      • OverlayOptions.get_callbacks()
      • OverlayOptions.load()
      • OverlayOptions.parse_added_filters()
      • OverlayOptions.update_filter_selector()
      • OverlayOptions.update_overlays()
      • OverlayOptions.update_slide()
    • PropertyPlotter
      • PropertyPlotter.extract_data_from_features()
      • PropertyPlotter.extract_property()
      • PropertyPlotter.gen_layout()
      • PropertyPlotter.gen_scatter_plot()
      • PropertyPlotter.gen_selected_div()
      • PropertyPlotter.gen_umap_cols()
      • PropertyPlotter.gen_violin_plot()
      • PropertyPlotter.generate_property_dict()
      • PropertyPlotter.get_callbacks()
      • PropertyPlotter.load()
      • PropertyPlotter.make_property_plot_tabs()
      • PropertyPlotter.remove_marker_label()
      • PropertyPlotter.select_data_from_plot()
      • PropertyPlotter.sub_select_data()
      • PropertyPlotter.update_property_graph()
      • PropertyPlotter.update_slide()
      • PropertyPlotter.update_sub_div()
    • PropertyViewer
      • PropertyViewer.gen_layout()
      • PropertyViewer.generate_plot_tabs()
      • PropertyViewer.get_callbacks()
      • PropertyViewer.load()
      • PropertyViewer.update_property_viewer()
      • PropertyViewer.update_slide()
    • Tool
• fusion_tools.utils package
  • fusion_tools.utils.shapes module
    • align_object_props()
    • convert_histomics()
    • detect_geojson()
    • detect_histomics()
    • export_annotations()
    • extract_geojson_properties()
    • extract_listed_prop()
    • extract_nested_prop()
    • find_intersecting()
    • find_nested_levels()
    • load_annotations()
    • load_aperio()
    • load_geojson()
    • load_histomics()
    • load_label_mask()
    • load_polygon_csv()
    • load_visium()
    • merge_dict()
    • path_to_indices()
    • path_to_mask()
    • process_filters_queries()
    • spatially_aggregate()
• fusion_tools.visualization package
  • fusion_tools.visualization.components module
    • Visualization
      • Visualization.check_for_uploader()
      • Visualization.gen_header_components()
      • Visualization.gen_layout()
      • Visualization.get_callbacks()
      • Visualization.get_layout_children()
      • Visualization.initialize_stores()
      • Visualization.open_header()
      • Visualization.open_header_component()
      • Visualization.open_navbar()
      • Visualization.start()
      • Visualization.update_page()
      • Visualization.update_page_layout()
  • fusion_tools.visualization.vis_utils module
    • get_pattern_matching_value()

fusion_tools.handler module

class fusion_tools.handler.DSATool

Bases: Tool

A sub-class of Tool specific to DSA components. The only difference is that these components always update relative to the session data.

Parameters:

Tool (None) – General class for components that perform visualization and analysis of data.

class fusion_tools.handler.Handler

Bases: object

fusion_tools.tileserver module

Tile server components

class fusion_tools.tileserver.CustomTileServer(tiles_url: str, regions_url: str, image_metadata: dict, annotations_url: str | None = None, name: str = None)

Bases: TileServer

CustomTileServer component if using some other tiles endpoint (must pass tileSize and levels in dictionary)

class fusion_tools.tileserver.DSATileServer(api_url: str, item_id: str, user_token: str | None = None)

Bases: TileServer

Use for linking visualization with remote tiles API (DSA server)

class fusion_tools.tileserver.LocalTileServer(local_image_path: str | list | None = [], local_image_annotations: str | list = [], tile_server_port: int = 8050, host: str = 'localhost')

Bases: TileServer

Tile server from image saved locally. Uses large-image to read and parse image formats (default: [common])

add_new_image(new_image_path: str, new_annotations: str | list | dict | None, new_metadata: dict | None = None)

get_annotations(image: int)

get_metadata(image: int)

Getting large-image metadata for image

Returns:

Dictionary containing metadata for local image

Return type:

Response

get_name_annotations_url(name)

get_name_metadata_url(name)

get_name_regions_url(name)

get_name_tiles_url(name)

get_names()

get_region(image: int, top: int, left: int, bottom: int, right: int, style: str = '')

Grabbing a specific region in the image based on bounding box coordinates

get_thumbnail(image: int)

Grabbing an image thumbnail

Parameters:

image (int) – _description_

get_tile(image: int, z: int, x: int, y: int, style: str = '')

Tiles endpoint, returns an image tyle based on provided coordinates

Parameters:

• z (int) – Zoom level for tile

• x (int) – X tile coordinate

• y (int) – Y tile coordinate

• style (dict, optional) – Additional style arguments to pass to large-image, defaults to {}

Returns:

Image tile containing bytes encoded pixel information

Return type:

Response

load_annotations()

root()

start()

Starting tile server instance on a provided port

Parameters:

port (str, optional) – Tile server port from which tiles are accessed, defaults to ‘8050’

class fusion_tools.tileserver.TileServer

Bases: object

Components which pull information from a slide(s)

fusion_tools.dataset module

Dataset sub-module for fusion-tools

class fusion_tools.dataset.ClassificationDataset(slides: list = [], annotations: list | dict | None = None, label_property: list | str | None = None, transforms: None | Callable = None, label_transforms: None | Callable = None, patch_size: list = [256, 256], patch_mode: str = 'bbox', patch_region: str | list | dict = 'all', spatial_filters: list | None = None, property_filters: list | None = None, use_structures: str | list = 'all', use_cache: bool = True, shuffle: bool = True, seed_val: int = 1701, verbose: bool = False, use_parallel: bool = True, n_jobs: int = 1)

Bases: object

export_configs(save_path: str)

gen_slide_patches()

get_next_image(idx: int)

make_image_and_label(bbox: list, slide_idx: int, features: list)

make_patch(bbox: list, slide_idx: int, annotations: GeoDataFrame)

plant_seed(seed_val)

process_annotations()

class fusion_tools.dataset.FUSIONDataset(slides: list = [], annotations: list | dict | None = None, image_transforms: Callable | None = None, target_transforms: Callable | None = None, patch_size: list = [256, 256], patch_mode: str = 'bbox', patch_region: str | list | dict = 'all', overlap: float | None = None, spatial_filters: list | None = None, property_filters: list | None = None, use_cache: bool = False, shuffle: bool = True, seed_val: int = 1701, verbose: bool = False, use_parallel: bool = True, n_jobs: int = 1)

Bases: object

find_tissue()

gen_slide_patches()

plant_seed(seed_val)

process_annotations()

class fusion_tools.dataset.SegmentationDataset(slides: list = [], annotations: list | dict | None = None, transforms: None | Callable = None, target_transforms: None | Callable = None, mask_key: dict | None = None, patch_size: list = [256, 256], patch_mode: str = 'bbox', patch_region: str | list | dict = 'all', spatial_filters: list | None = None, property_filters: list | None = None, use_structures: str | list = 'all', use_cache: bool = True, shuffle: bool = True, seed_val: int = 1701, verbose: bool = False, use_parallel: bool = True, n_jobs: int = 1)

Bases: object

export_configs(save_path: str)

find_tissue(slide_tile_source, slide_metadata)

gen_slide_patches()

get_next_image(idx: int)

make_image_and_mask(bbox: list, slide_idx: int, features: list)

make_patch(bbox: list, slide_idx: int, annotations: GeoDataFrame)

plant_seed(seed_val)

process_annotations()

fusion_tools.feature_extraction module

Feature extraction functions

class fusion_tools.feature_extraction.ParallelFeatureExtractor(image_source: str | TileServer = None, feature_list: list = [], preprocess: Callable | None = None, sub_mask: Callable | None = None, mask_names: list | None = None, channel_names: list | None = None, n_jobs: None = None, verbose: bool = False)

Bases: object

This class is for extracting multiple types of features at the same time using joblib

extract_features(region: dict) → dict

get_bbox(coords: list) → list

make_mask(coords: list) → ndarray

merge_dict(a: dict, b: dict, path=[])

read_image_region(coords: list) → ndarray

start(region_list: list) → DataFrame

update_dict_keys(input_dict: dict)

fusion_tools.feature_extraction.color_features(image: ndarray, mask: ndarray, coords: list) → dict

Calculate “color” features for each label in mask within image (color defined as channel statistics)

Parameters:

• image (np.ndarray) – Input image region

• mask (np.ndarray) – Mask of regions to include in the feature calculation (0=background)

• coords (list) – Coordinates of this specific image (not used)

Returns:

Dictionary containing key/value pairs for each feature extracted (Mean, Median, Maximum, Std)

Return type:

dict

fusion_tools.feature_extraction.distance_transform_features(image: ndarray, mask: ndarray, coords: list) → dict

Function to calculate distance transform features for each label in “mask”

Parameters:

• image (np.ndarray) – Input image region (not used)

• mask (np.ndarray) – Mask of regions to include in the feature calculation (0=background)

• coords (list) – Coordinates of this specific image (not used)

Returns:

Dictionary containing key/value pairs for each feature extracted (Mean, Median, Maximum, Sum)

Return type:

dict

fusion_tools.feature_extraction.morphological_features(image: ndarray, mask: ndarray, coords: list) → dict

Calculate morphological features for each label in mask within image.

Parameters:

• image (np.ndarray) – Input image region (not used)

• mask (np.ndarray) – Mask of regions to include in the feature calculation (0=background)

• coords (list) – Coordinates of this specific image (not used)

Returns:

Dictionary containing key/value pairs for each feature extracted (Contrast, Homogeneity, Correlation, Energy)

Return type:

dict

fusion_tools.feature_extraction.relative_distance(input_shapes: dict, other_shapes: dict | list) → list

Calculate relative distance statistics between each Feature in “input_shapes” (GeoJSON FeatureCollection) and each Feature in each FeatureCollection in “other_shapes”

Parameters:

• input_shapes (dict) – FeatureCollection containing Features to calculate relative distance statistics for each FeatureCollection in other_shapes

• other_shapes (Union[dict,list]) – List of multiple FeatureCollections or single FeatureCollection where relative distance statistics are calculated off of for input_shapes

fusion_tools.feature_extraction.texture_features(image: ndarray, mask: ndarray, coords: list) → dict

Calculate texture features for each label in mask within image.

Parameters:

• image (np.ndarray) – Input image region

• mask (np.ndarray) – Mask of regions to include in the feature calculation (0=background)

• coords (list) – Coordinates of this specific image (not used)

Returns:

Dictionary containing key/value pairs for each feature extracted (Contrast, Homogeneity, Correlation, Energy)

Return type:

dict

fusion_tools.feature_extraction.threshold_channels(input_image: ndarray, threshold_method: str | None = None) → ndarray

Example preprocessing function that thresholds each channel in the input image according to some method

Parameters:

• input_image (np.ndarray) – Input image region (Y,X,C)

• threshold_method (Union[str,None], optional) – Method to use for thresholding (None is set to Otsu’s), defaults to None

Returns:

Returns image with each channel thresholded independently

Return type:

np.ndarray