Components Overview

Here you’ll find documentation for each of the individual components and an overview on their usage.

Map Components

These components control how a slide and associated annotations are viewed.

Slide Map

MultiFrame Slide Map

Overlay Options

Channel Mixer

Slide Image Overlay

Components used for high-resolution image viewing

class fusion_tools.components.maps.ChannelMixer

Bases: MapComponent

ChannelMixer component that allows users to select various frames from their image to overlay at the same time with different color (styles) applied.

Parameters:

MapComponent (None) – General component class for children of SlideMap

add_color_selector_tab(channel_mix_values: list | None, current_channels: list | None, current_colors: list | None)

Add a new color selector tab to channel-mixer-parent for selecting overlaid channel color

Parameters:

  • channel_mix_values (list) – Selected list of channels to overlay

  • current_channels (list) – Current set of overlaid channels (names)

  • current_colors (list) – Current set of overlaid channels (colors)

gen_layout(session_data: dict)

Generating layout for ChannelMixer component

get_callbacks()

Initializing callbacks and adding to DashBlueprint

load(component_prefix: int)
process_frames(image_metadata: dict)

Extracting names for each frame for easy reference

updat_layout(session_data: dict, use_prefix: bool)

Updating layout of ChannelMixer component

Parameters:

  • session_data (dict) – Current data relating to visualization session

  • use_prefix (bool) – Whether or not this is the initial load of the component

Returns:

ChannelMixer component

update_channel_mix(butt_click: list, current_channels: list, current_colors: list, frame_names: list, slide_info: list)

Updating urls of all tile layers to include selected overlay channels

Parameters:

  • butt_click (list) – Button clicked to update channel mix

  • current_channels (list) – Names of channels to overlay

  • current_colors (list) – Colors of channels to overlay

update_slide(selected_slide, vis_data)

Updating component data when a new slide is selected

Parameters:

  • selected_slide (list) – New slide

  • vis_data (str) – Data relating to current visualization session

update_tab_style(color_select)

Updating color of tab label based on selection

Parameters:

color_select (str) – “rgba” formatted color selection

class fusion_tools.components.maps.MapComponent

Bases: object

General class for components added to SlideMap For more information see dash-leaflet: https://www.dash-leaflet.com/

class fusion_tools.components.maps.MultiFrameSlideMap

Bases: SlideMap

MultiFrameSlideMap component, containing an image with multiple frames which are added as additional, selectable dl.TileLayer() components

Parameters:

SlideMap (None) – dl.Map() container where image tiles are displayed.

gen_layout(session_data: dict)

Generating layout for MultiFrameSlideMap

load(component_prefix: int)
process_frames(image_metadata, tiles_url)

Create BaseLayer and TileLayer components for each of the different frames present in a multi-frame image Also initializes base tile url and styled tile url

update_layout(session_data: dict, use_prefix: bool)

Updating layout of MultiFramSlideMap component

Parameters:

  • session_data (dict) – Data relating to current visualization session

  • use_prefix (bool) – Whether or not this is the initial loading of the component

Returns:

Component layout

class fusion_tools.components.maps.SlideImageOverlay(image_path: str, image_crs: list = [0, 0], image_properties: dict | None = {'None': ''})

Bases: MapComponent

Image overlay on specific coordinates within a SlideMap

Parameters:

MapComponent (None) – General component class for children of SlideMap

get_image_bounds()

Get total bounds of image overlay in original CRS (number of pixels)

Returns:

List of image bounds in overlay CRS ([minx, miny, maxx, maxy])

Return type:

list

to_dict()
class fusion_tools.components.maps.SlideMap

Bases: MapComponent

This is a general high-resolution tiled image component.

Parameters:

MapComponent (None) – General class for components added to SlideMap

add_manual_roi(new_geojson: list, uploaded_shape: list, current_annotations: list, frame_layers: list, line_colors: list, colormap: list, separate_switch: list, summarize_switch: list) → list

Adding a manual region of interest (ROI) to the SlideMap using dl.EditControl() tools including polygon, rectangle, and markers.

Parameters:

  • new_geojson (list) – Incoming GeoJSON object that is emitted by dl.EditControl() following annotation on SlideMap

  • current_annotations (list) – Initial annotations added to the map on initialization

  • annotation_names (list) – Names for each annotation layer on the slide map

  • frame_layers (list) – Frame layers for multi-frame visualization

Raises:

  • exceptions.PreventUpdate – new_geojson input is None

  • exceptions.PreventUpdate – No new features are added, this can occur after deletion of previous manual ROIs.

Returns:

List of new children to dl.LayerControl() consisting of overlaid GeoJSON components.

Return type:

list

download_manual_roi(button_click, current_manual_geojson, slide_information)
export_image_overlay_bounds(button_click, current_position, slide_information)

Exporting position of image overlay

Parameters:

  • button_click (int) – Export button clicked

  • current_position (list) – Current position of the image overlay

Raises:

exceptions.PreventUpdate – Preventing callback if button isn’t clicked

Returns:

JSON file containing original image path and current position on the slide

Return type:

dict

gen_image_mover(button_click, current_bounds, button_text)

Creates a draggable Marker to facilitate moving an image overlay around

Parameters:

  • button_click (dict) – Position button is clicked

  • current_bounds (list) – Current bounds of the image overlay, used for setting initial position of mover Marker

  • button_text (str) – Text displayed on position button.

Raises:

exceptions.PreventUpdate – Prevent update if button is not pressed.

Returns:

New text for the position button and mover Marker

Return type:

tuple

gen_layout(session_data: dict)
get_callbacks()

Initializing callback functions for interactive components in SlideMap. Adding these callbacks to the DashBlueprint() object enable embedding into other layouts.

get_click_popup(clicked)

Populating popup Div with summary information on the clicked GeoJSON feature

Parameters:

clicked (dl.GeoJSON.GeoJSON) – GeoJSON feature on SlideMap that was selected

Raises:

exceptions.PreventUpdate – Stops callback execution

Returns:

Popup with summary information on clicked GeoJSON. Nested properties (in dictionaries) appear as collapsible dbc.AccordionItem() with their own dash_table.DataTable

Return type:

dash.html.Div.Div

get_image_overlay_popup(st, st_idx)

Getting popup components for image overlay annotations

Parameters:

  • st (SlideImageOverlay) – New ImageOverlay annotation

  • st_idx (int) – Index to use for interactive components

get_namespace()

Adding JavaScript functions to the SlideMap Namespace

These functions are automatically written to .fusion_assets/fusionTools_default.js where they are accessible to select interactive components in SlideMap.

For more information on adding JS functions to a Dash application, see:

https://www.dash-extensions.com/sections/javascript and https://www.dash-leaflet.com/docs/geojson_tutorial

get_scale_factors(image_metadata: dict)

Function used to initialize scaling factors applied to GeoJSON annotations to project annotations into the SlideMap CRS (coordinate reference system)

Returns:

x and y (horizontal and vertical) scale factors applied to each coordinate in incoming annotations

Return type:

float

load(component_prefix: int)
make_geojson_layers(geojson_list: list, names_list: list, index_list: list, line_color: list, colormap: str) → list

Creates new dl.Overlay() dl.GeoJSON components from list of GeoJSON FeatureCollection objects

Parameters:

  • geojson_list (list) – List of GeoJSON FeatureCollection objects

  • names_list (list) – List of names for each layer

  • index_list (list) – List of indices to apply to each component

Returns:

Overlay components on SlideMap.

Return type:

list

move_image_overlay(new_position, old_bounds)

Move image overlay to location of mover Marker

Parameters:

  • new_position (dict) – New top-left position of image overlay (keys: lat, lng)

  • old_bounds (list) – Old bounds of the image, used for determining bottom-right coordinates

Raises:

exceptions.PreventUpdate – Prevents callback execution until a position is set

Returns:

New bounds for image overlay component

Return type:

list

update_image_overlay_transparency(new_opacity: float)

Update transparency of image overlay component

Parameters:

new_opacity (float) – New opacity level (0-1 (0.1 increments)) from slider

Returns:

New opacity level

Return type:

float

update_layout(session_data: dict, use_prefix: bool)

Generating SlideMap layout

Returns:

Div object containing interactive components for the SlideMap object.

Return type:

dash.html.Div.Div

update_slide(slide_selected, vis_data)
update_vis_session(new_vis_data)

Updating slide dropdown options based on current visualization session

Parameters:

new_vis_data (str) – Visualization session data containing information on selectable slides

Returns:

New options for slide dropdown

Return type:

list

upload_shape(upload_clicked, is_open)

Plotting Components

Visualization tools which can be linked to SlideMap components (but don’t have to be)

class fusion_tools.components.tools.GlobalPropertyPlotter(ignore_list: list = [], property_depth: int = 6, preloaded_properties: DataFrame | None = None, structure_column: str | None = None, slide_column: str | None = None, bbox_columns: list | str | None = None, nested_prop_sep: str = ' --> ')

Bases: MultiTool

extract_all_properties(session_data)
extract_property_info()
gen_layout(session_data: dict)
gen_scatter_plot(data_df: DataFrame, plot_cols: list, label_col: str | None, customdata_cols: list)

Generating a 2D scatter plot using provided data

Parameters:

  • data_df (pd.DataFrame) – Extracted data from current GeoJSON features

  • plot_cols (list) – Names of columns containing properties for the scatter plot

  • label_cols (Union[str,None]) – Name of column to use to label markers.

  • customdata_cols (list) – Names of columns to use for customdata on plot

Returns:

Scatter plot figure

Return type:

go.Figure

gen_umap_cols(data_df: DataFrame, property_columns: list) → DataFrame

Scale and run UMAP for dimensionality reduction

Parameters:

  • data_df (pd.DataFrame) – Extracted data from current GeoJSON features

  • property_columns (list) – Names of columns containing properties for UMAP plot

Returns:

Dataframe containing colunns named UMAP1 and UMAP2

Return type:

pd.DataFrame

gen_violin_plot(data_df: DataFrame, label_col: str | None, property_column: str, customdata_columns: list)

Generating a violin plot using provided data, property columns, and customdata

Parameters:

  • data_df (pd.DataFrame) – Extracted data from current GeoJSON features

  • label_col (Union[str,None]) – Name of column to use for label

  • property_column (str) – Name of column to use for property (y-axis) value

  • customdata_columns (list) – Names of columns to use for customdata (accessible from clickData and selectedData)

Returns:

Figure containing violin plot data

Return type:

go.Figure

generate_plot(butt_click, current_data)

Generate a new plot based on selected properties

Parameters:

  • butt_click (list) – Generate Plot button clicked

  • current_data (list) – Dictionary containing current selected properties (in a list)

generate_property_dict(available_properties, title: str = 'Properties')
get_callbacks()
load(component_prefix: int)
parse_filter_divs(add_property_parent: list) → list

Processing parent div object and extracting name and range for each property filter.

Parameters:

add_property_parent (list) – Parent div containing property filters

Returns:

List of property filters (dicts with keys: name and range)

Return type:

list

select_data_from_plot(selected_data, session_data)

Select point(s) from the plot and extract the image from that/those point(s)

Parameters:

  • selected_data (list) – Multiple points selected using either a box or lasso select

  • session_data (list) – Data for each slide in the current session

update_drop_type(switch_switched)

Update the property selection mode (either dropdown menu or tree view)

Parameters:

switch_switched (list) – Switch selected

Returns:

New property selector component

Return type:

list

update_filter_properties(add_click: list, remove_click: list, property_info: list)

Adding/removing property filter dropdown

Parameters:

  • add_click (list) – Add property filter clicked

  • remove_click (list) – Remove property filter clicked

Returns:

Property filter dropdown and parent div of range selector

Return type:

tuple

update_properties_and_filters(property_selection, property_checked, structure_selection, label_selection, filter_prop_mod, filter_prop_remove, filter_prop_selector, property_divs, current_data)

Updating the properties, structures, and filters incorporated into the main plot

Parameters:

  • property_selection (list) – Properties selected for plotting

  • property_checked (list) – Properties selected from property tree view

  • structure_selection (list) – Structures selected to be plotted

  • label_selection (list) – Label selected for the plot

  • remove_prop (list) – Remove property filter clicked

  • prop_mod (list) – Modifier applied to property filter changed

  • property_divs (list) – Children of property filter parent has been updated

  • current_data (list) – Contents of the global-property-store, updated by this plugin.

Returns:

Updated global-property-store object

Return type:

list

update_property_selector(property_value: str, property_info: list)

Updating property filter range selector

Parameters:

  • property_value (str) – Name of property to generate selector for

  • property_info (list) – Dictionary containing range/unique values for each property

Returns:

Either a multi-dropdown for categorical properties or a RangeSlider for quantitative values

Return type:

tuple

class fusion_tools.components.tools.HRAViewer

Bases: Tool

HRAViewer Tool which enables hierarchy visualization for organs, cell types, biomarkers, and proteins in the Human Reference Atlas

For more information on the Human Reference Atlas (HRA), see: https://humanatlas.io/

Parameters:

Tool (None) – General class for interactive components that visualize, edit, or perform analyses on data.

gen_layout(session_data: dict)

Generate layout for HRA Viewer component

get_callbacks()

Initializing callbacks and attaching to DashBlueprint

get_organ_table(organ: str)

Grabbing ASCT+B Table for a specific organ

Parameters:

organ (str) – Name of organ to get table for

load(component_prefix: int)
update_hra_viewer(viewer_drop_value)

Updating the HRAViewer component based on selected view

Parameters:

viewer_drop_value (list) – Selected component from dropdown (one of FTU Explorer or {organ} ASCT+B Table)

class fusion_tools.components.tools.MultiTool

Bases: object

General class for a Tool which works on multiple slides at once

class fusion_tools.components.tools.OverlayOptions(ignore_list: list = ['_id', '_index'], property_depth: int = 4)

Bases: Tool

OverlayOptions Tool which enables editing overlay visualization properties including line color, fill color, and filters.

Parameters:

Tool (None) – General class for interactive components that visualize, edit, or perform analyses on data.

add_filter(add_filter_click, delete_filter_click, overlay_info_state, overlay_options)

Adding a new filter to apply to GeoJSON features

Parameters:

  • add_filter_click (_type_) – Add filter icon clicked

  • delete_filter_click (_type_) – Delete filter icon (x) clicked

  • overlay_info_state (_type_) – Summary information on properties for GeoJSON features

Raises:

  • exceptions.PreventUpdate – Stop callback execution

  • exceptions.PreventUpdate – No information provided for GeoJSON properties, cancels adding new filter.

Returns:

List of current filters applied to GeoJSON features

Return type:

list

add_structure_line_color(overlay_names: list)
adv_update_overlays(butt_click: list, colorbar_width: list, colormap_val: list, current_feature_hideout: list)

Update some additional properties of the overlays and display.

Parameters:

  • butt_click (list) – Button clicked to update overlay properties

  • colorbar_width (list) – Width of colorbar in pixels

  • colormap_val (list) – Colormap option to pass to chroma

  • current_feature_hideout (list) – Current hideout properties for all structures

Returns:

Updated colorbar width, line width, and colormap

Return type:

tuple

create_layer_from_filter(icon_click: list, current_filters: list, current_annotations: list, current_overlay_names: list, line_colors: list, colormap: list)
export_layers(button_click, current_layers, slide_information)
gen_layout(session_data: dict)

Generating OverlayOptions layout, added to DashBlueprint() object to be embedded in larger layout.

get_callbacks()

Initializing callbacks for OverlayOptions Tool

load(component_prefix: int)
parse_added_filters(add_filter_parent: list) → list

Getting all filter values from parent div

Parameters:

add_filter_parent (list) – List of dictionaries containing component information from dynamically added filters.

Returns:

List of filter values to apply to current GeoJSON features.

Return type:

list

update_filter_selector(add_filter_value, overlay_info_state)

Updating the filter selector (either a RangeSlider or Dropdown component depending on property type)

Parameters:

  • add_filter_value (str) – Selected property to be used for filters.

  • overlay_info_state (list) – Current information on GeoJSON feature properties (contains “min” and “max” for numeric properties and “unique” for categorical properties)

Raises:

exceptions.PreventUpdate – Stop callback execution

Returns:

Filter selector. Either a dropdown menu for categorical properties or a dcc.RangeSlider for selecting a range of values between the minimum and maximum for that property

Return type:

_type_

update_overlays(overlay_value, transp_value, lineColor_butt, filter_parent, filter_value, delete_filter, overlay_state, transp_state, overlay_info_state, lineColor_state, overlay_names, colormap_val, current_hideout)

Update overlay transparency and color based on property selection

Adding new values to the “hideout” property of the GeoJSON layers triggers the featureStyle Namespace function

Parameters:

  • overlay_value (list) – Value to use as a basis for generating colors (range of colors for numeric properties and list of colors for categorical properties)

  • transp_value (list) – Transparency value for each feature.

  • lineColor_butt (list) – Whether the button to update the lineColor property was clicked.

  • filter_parent (list) – Parent Div containing filter information

  • filter_value (list) – Whether the callback is triggered by a new filter value being selected

  • delete_filter (list) – Whether the callback is triggered by a filter being deleted

  • overlay_state (list) – State of overlay dropdown if not triggered by new overlay value selection

  • transp_state (list) – State of transparency slider if not triggered by transparency slider adjustment

  • overlay_info_state (list) – Information on overlay properties for current GeoJSON features

  • lineColor_state (list) – Current selected lineColors to apply to current GeoJSON features

  • colormap_val (list) – Current colormap applied to overlays,

  • current_hideout (list) – Current hideout properties assigned to each GeoJSON layer

Returns:

List of dictionaries added to the GeoJSONs’ “hideout” property (used by Namespace functions) and a colorbar based on overlay value.

Return type:

tuple

update_slide(new_annotations: list)
class fusion_tools.components.tools.PropertyPlotter(ignore_list: list = [], property_depth: int = 6)

Bases: Tool

PropertyPlotter Tool which enables more detailed selection of properties across the entire tissue. Allows for generation of violin plots, scatter plots, and UMAP plots.

Parameters:

Tool (None) – General class for interactive components that visualize, edit, or perform analyses on data.

extract_data_from_features(geo_list: list, properties: list, labels: list, filter_list: list | None = None) → list

Iterate through properties and extract data based on selection

Parameters:

  • geo_list (list) – List of current GeoJSON features

  • properties (list) – List of properties to use in the current plot

  • labels (list) – List of labels to use in the current plot (should just be one element)

  • filter_list – List of dictionaries containing g_idx and f_idx corresponding to the layers and features to extract data from

  • filter_list – Union[list,None], optional

Returns:

Extracted data to use for the plot

Return type:

list

extract_property(feature: dict, properties: list, labels: list) → dict

Extracting list of properties and labels from a single feature in a GeoJSON FeatureCollection

Parameters:

  • feature (dict) – One Feature in GeoJSON FeatureCollection

  • properties (list) – List of property names to extract

  • labels (list) – List of properties used as labels to extract

Returns:

Dictionary of properties and labels for a single Feature

Return type:

dict

gen_layout(session_data: dict)

Generating layout for PropertyPlotter Tool

Returns:

Layout for PropertyPlotter DashBlueprint object to be embedded in larger layouts

Return type:

dash.html.Div.Div

gen_scatter_plot(data_df: DataFrame, plot_cols: list, label_col: str | None, customdata_cols: list)

Generating a 2D scatter plot using provided data

Parameters:

  • data_df (pd.DataFrame) – Extracted data from current GeoJSON features

  • plot_cols (list) – Names of columns containing properties for the scatter plot

  • label_cols (Union[str,None]) – Name of column to use to label markers.

  • customdata_cols (list) – Names of columns to use for customdata on plot

Returns:

Scatter plot figure

Return type:

go.Figure

gen_selected_div(n_markers: int, available_properties)

Generate a new property-graph-selected-div after changing the number of markers

Parameters:

n_markers (int) – New number of markers

gen_umap_cols(data_df: DataFrame, property_columns: list) → DataFrame

Scale and run UMAP for dimensionality reduction

Parameters:

  • data_df (pd.DataFrame) – Extracted data from current GeoJSON features

  • property_columns (list) – Names of columns containing properties for UMAP plot

Returns:

Dataframe containing colunns named UMAP1 and UMAP2

Return type:

pd.DataFrame

gen_violin_plot(data_df: DataFrame, label_col: str | None, property_column: str, customdata_columns: list)

Generating a violin plot using provided data, property columns, and customdata

Parameters:

  • data_df (pd.DataFrame) – Extracted data from current GeoJSON features

  • label_col (Union[str,None]) – Name of column to use for label

  • property_column (str) – Name of column to use for property (y-axis) value

  • customdata_columns (list) – Names of columns to use for customdata (accessible from clickData and selectedData)

Returns:

Figure containing violin plot data

Return type:

go.Figure

generate_property_dict(available_properties, title: str = 'Features')
get_callbacks()

Initializing callbacks for PropertyPlotter Tool

load(component_prefix: int)
make_property_plot_tabs(data_df: DataFrame, label_col: str | None, property_cols: list, customdata_cols: list) → list

Generate property plot description tabs

Parameters:

  • data_df (pd.DataFrame) – Current data used to generate the plot

  • label_col (Union[str,None]) – Column of data_df used to label points in the plot.

  • property_cols (list) – Column(s) of data_df plotted in the plot.

  • customdata_cols (list) – Column(s) used in the “customdata” field of points in the plot

Returns:

List of tabs including summary statistics, clustering options, selected data options

Return type:

list

remove_marker_label(delete_click, current_plot_data, available_properties)

Remove marker from selected feature

Parameters:

  • delete_click (list) – Clear marker button clicked

  • current_plot_data (list) – Current data for plot

Returns:

Updating current markers on SlideMap, current plot data, and property-graph-selected-div

Return type:

tuple

select_data_from_plot(selected_data, current_plot_data, available_properties)

Updating selected data tab based on selection in primary graph

Parameters:

  • selected_data (list) – Selected data points in the current plot

  • current_plot_data (list) – Data used to generate the current plot

Raises:

  • exceptions.PreventUpdate – Stop callback execution because there is no selected data

  • exceptions.PreventUpdate – Stop callback execution because there is no selected data

Returns:

Selected data description div, markers to add to the map, updated current plot data

Return type:

tuple

sub_select_data(sub_selected_data)

Updating the markers on the map according to selected data in the sub-plot

Parameters:

  • sub_selected_data (list) – Selected points in the sub-plot

  • current_plot_data (list) – Plot data in the main-plot

  • current_features (list) – Current GeoJSON features on the map

Returns:

Updated list of markers applied to the map

Return type:

list

update_property_graph(plot_butt_click, property_list, label_list, property_keys, current_features, current_plot_data)

Updating plot based on selected properties, label(s)

Parameters:

  • plot_butt_click (list) – Whether this callback is triggered by the plot button being clicked

  • property_list (list) – List of properties to incorporate in the generated plot

  • label_list (list) – List of properties to use as labels in the generated plot

  • current_features (list) – Current set of GeoJSON features

  • current_plot_data (list) – Data present in the current plot.

Raises:

  • exceptions.PreventUpdate – Stop callback execution because no properties are selected

  • exceptions.PreventUpdate – Stop callback execution because no properties are selected

Returns:

Generated plot figure and new plot data

Return type:

tuple

update_slide(new_annotations: list)
update_sub_div(plot_butt_clicked, marker_butt_clicked, sub_plot_value, current_plot_data, current_features, current_labels)

Updating the property-graph-selected-div based on selection of either a property to plot a sub-plot of or whether the marker properties button was clicked

Parameters:

  • plot_butt_clicked (list) – Update sub-plot button was clicked

  • marker_butt_clicked (list) – Get marker features for selected samples clicked

  • current_plot_data (list) – Current data in the plot

  • current_features (list) – Current set of GeoJSON FeatureCollections on the SlideMap

  • current_labels (list) – Current labels applied to the main plot

Returns:

Updated children of the selected-property-graph-div including either sub-plot or table of marker features

Return type:

tuple

class fusion_tools.components.tools.PropertyViewer(ignore_list: list = [], property_depth: int = 6)

Bases: Tool

PropertyViewer Tool which allows users to view distribution of properties across the current viewport of the SlideMap

Parameters:

Tool (None) – General class for interactive components that visualize, edit, or perform analyses on data

gen_layout(session_data: dict)

Generating layout for PropertyViewer Tool

Returns:

Layout added to DashBlueprint() object to be embedded in larger layout

Return type:

dash.html.Div.Div

generate_plot_tabs(current_property_data, current_features)

Generating tabs for each structure containing plot of selected value.

Parameters:

  • current_property_data (dict) – Data stored for current viewport and property value selection

  • current_features (list) – Current GeoJSON features

Returns:

Tabs object containing a tab for each structure with a plot of the selected property value

Return type:

dbc.Tabs

get_callbacks()

Initializing callbacks for PropertyViewer Tool

load(component_prefix: int)
update_property_viewer(slide_map_bounds, view_type_value, view_subtype_value, view_butt_click, active_tab, current_property_data, update_viewer, current_subtype_children, current_geojson, available_properties)

Updating visualization of properties within the current viewport

Parameters:

  • slide_map_bounds (list) – Current viewport boundaries

  • view_type_value (list) – Property to view across each GeoJSON object

  • view_subtype_value (list) – If a subtype is available for the view type, what it’s value is. This is present for nested (dictionary) properties.

  • view_butt_click (list) – Whether the callback is triggered by update plot button

  • active_tab (list) – Which tool tab is currently in view. Prevents update if the current tab isn’t “property-viewer”

  • current_property_data (list) – Data used in current set of plots

  • update_viewer (list) – Switch value for whether or not to update the plots based on panning around in the SlideMap

  • current_subtype_children (list) – Parent container of all sub-type dropdown divs.

  • current_geojson (list) – Current set of GeoJSON features and their properties

Raises:

exceptions.PreventUpdate – Stop callback execution

Returns:

List of PropertyViewer tabs (separated by structure) and data used in plots

Return type:

tuple

update_slide(new_annotations: list)
class fusion_tools.components.tools.Tool

Bases: object

General class for interactive components that visualize, edit, or perform analyses on data.

Annotation Components

Components related to segmentation

class fusion_tools.components.segmentation.BulkLabels(ignore_list: list = [], property_depth: int = 4)

Bases: Tool

Add labels to many structures at the same time

Parameters:

Tool (None) – General class for interactive components that visualize, edit, or perform analyses on data

add_labels_to_structures(button_click, current_annotations, current_labels)
apply_labels(button_click, current_markers, current_data, label_type, label_text, label_rationale, label_source, label_method, slide_information)

Applying current label to marked structures

Parameters:

  • button_click (list) – Button clicked

  • current_markers (list) – Current markers on the slide-map

  • current_data (list) – Current labels data

  • label_type (list) – Type of label being added

  • label_text (list) – Label being added to structures

  • label_rationale (list) – Optional additional rationale to save with label decision

  • label_source (list) – Property and spatial filters incorporated to generate markers

  • label_method (list) – Method used to apply labels to structures

  • current_annotations (list) – Current GeoJSON annotations on the slide map

Returns:

Updated labels data and annotations with label added as a property

Return type:

tuple

download_data(button_click, label_data)

Download both the labels applied to the current session as well as the labeling metadata

Parameters:

  • button_click (list) – Download button clicked

  • label_data (list) – Current label data and metadata

Returns:

Data sent to the two download components in the layout containing JSON formatted label data and metadata

Return type:

tuple

gen_layout(session_data: dict)

Generating layout for BulkLabels component

Returns:

BulkLabels layout

Return type:

html.Div

get_callbacks()

Adding callbacks to DashBlueprint object

get_namespace()

Adding JavaScript functions to the BulkLabels Namespace

get_scale_factors(image_metadata: dict)

Function used to initialize scaling factors applied to GeoJSON annotations to project annotations into the SlideMap CRS (coordinate reference system)

Returns:

x and y (horizontal and vertical) scale factors applied to each coordinate in incoming annotations

Return type:

float

load(component_prefix: int)
parse_filter_divs(add_property_parent: list) → list

Processing parent div object and extracting name and range for each property filter.

Parameters:

add_property_parent (list) – Parent div containing property filters

Returns:

List of property filters (dicts with keys: name and range)

Return type:

list

parse_spatial_divs(spatial_query_parent: list, slide_information: dict) → list

Parsing through parent div containing all spatial queries and returning them in list form

Parameters:

spatial_query_parent (list) – Div containing spatial query child divs

Returns:

Processed queries with keys “type”, “structure”, and “distance” (if “nearest” selected)

Return type:

list

refresh_labels(refresh_click, structure_options)

Clear current label components and start over

Parameters:

  • refresh_click (list) – Refresh icon clicked

  • structure_options (list) – Names of current overlay layers

Returns:

Cleared labeling components

Return type:

tuple

search_labels(marker_features, current_labels, label_type, label_text, label_method)

Checking current labels and adding the new label based on label_method (inclusive, exclusive, or unlabeled).

Parameters:

  • marker_features (list) – GeoJSON Features for new markers

  • current_labels (list) – List of all current labels (dicts with keys: centroid, labels)

  • label_type (str) – String corresponding to the type for the new label

  • label_text (str) – String corresponding to the new label to apply to each centroid

  • label_method (str) – One of: in (inclusive, append new label to current regardless of type), over (exclusive, replace existing labels of same type), or un (unlabeled, only apply label to structures that do not have a label for this type)

Returns:

Updated set of labels, including original set

Return type:

dict

update_filter_data(property_filter, sp_query_type, sp_query_structure, sp_query_distance, remove_sq, remove_prop, spatial_mod, prop_mod, property_divs: list, spatial_divs: list, slide_information: list)
update_label_properties(add_click: list, remove_click: list, property_info: list)

Adding/removing property filter dropdown

Parameters:

  • add_click (list) – Add property filter clicked

  • remove_click (list) – Remove property filter clicked

Returns:

Property filter dropdown and parent div of range selector

Return type:

tuple

update_label_structures(update_structures_click: list, include_structures: list, filter_data: list, current_features: list, slide_information: list)

Go through current structures and return all those that pass spatial and property filters.

Parameters:

  • update_structures_click (list) – Button clicked

  • include_structures (list) – List of structures to include in the final filtered GeoJSON

  • filter_properties (list) – List of property filters to apply to current GeoJSON

  • spatial_queries (list) – List of spatial filters to apply to current GeoJSON

  • current_features (list) – Current structures in slide map

Returns:

Markers over structures that pass the spatial and property filters, count of structures, enabled label components

Return type:

tuple

update_label_table(table_data, current_labels)
update_method_explanation(method)

Updating explanation given for the selected label method

Parameters:

method (list) – Name of method selected

Returns:

Description of the labeling method selected

Return type:

list

update_property_selector(property_value: str, property_info: list)

Updating property filter range selector

Parameters:

  • property_value (str) – Name of property to generate selector for

  • property_info (list) – Dictionary containing range/unique values for each property

Returns:

Either a multi-dropdown for categorical properties or a RangeSlider for quantitative values

Return type:

tuple

update_slide(new_annotations: list)
update_spatial_queries(add_click, remove_click, structure_names)

Generating a new spatial query selector once the add/remove icons are clicked

Parameters:

  • add_click (list) – Add spatial query icon selected

  • remove_click (list) – Delete spatial query icon selected

  • structure_names (list) – Names of current overlay structures

Returns:

Spatial query selectors (two dropdown menus and delete icon)

Return type:

tuple

update_spatial_query_definition(query_type)

Returning the definition for the selected spatial predicate type

Parameters:

query_type (str) – Name of spatial predicate

Returns:

Markdown string

Return type:

dcc.Markdown

update_structure_count(marker_dblclicked, marker_geo)
class fusion_tools.components.segmentation.FeatureAnnotation(storage_path: str, labels_format: str = 'json', annotations_format: str = 'one-hot')

Bases: Tool

Enables annotation (drawing) on top of structures in the SlideMap using a separate interface.

Parameters:

Tool (None) – General Class for interactive components that visualize, edit, or perform analyses on data.

add_new_class_label(add_class, add_label, new_class, new_label, label_text, current_structure_data, current_structure, slide_information)

Adding a new class or label to the current structure

Parameters:

  • add_class (list) – Add class clicked

  • add_label (list) – Add label clicked

  • new_class (list) – New class to add to the current structure

  • new_label (list) – New label name to add for the current structure

  • label_text (list) – New label text to add for the current structure

  • current_structure_data (list) – Current structure bounding boxes and indices

  • current_structure (list) – Current structure selection

Returns:

Returning updated figure with new line color, fill color and saving the new label to the save directory

Return type:

list

create_class_label(add_class_value, add_submit_click, add_class_color, add_class_name, add_label_name, current_class_options, current_label_options)

Creating a new class or label to add to the session

Parameters:

  • add_class_value (list) – Dropdown value for adding either a class or label

  • add_submit_click (list) – Submit button clicked

  • add_class_color (list) – New color for the new class added

  • add_class_name (list) – New name for the new class added

  • add_label_name (list) – New name for new label added

  • current_class_options (list) – Current options in the class dropdown menu

  • current_label_options (list) – Current options in the label dropdown menu

Returns:

Updating the label and class dropdown options

Return type:

tuple

gen_layout(session_data: dict)

Generating layout for component

get_callbacks()

Initializing callbacks and adding to DashBlueprint

get_scale_factors(image_metadata: dict)

Function used to initialize scaling factors applied to GeoJSON annotations to project annotations into the SlideMap CRS (coordinate reference system)

Returns:

x and y (horizontal and vertical) scale factors applied to each coordinate in incoming annotations

Return type:

float

get_structure_region(structure_bbox: list, slide_information: dict, scale: bool = True)

Using the tile server “regions_url” property to pull out a specific region of tissue

Parameters:

  • structure_bbox (list) – List of minx, miny, maxx, maxy coordinates

  • scale (bool) – Whether or not to scale the bbox values to slide coordinates

load(component_prefix: int)
save_annotation(save_click, current_annotations, current_classes, current_structure_data, current_structure, slide_information)

Saving the current annotation in image format

Parameters:

  • save_click (list) – Save button is clicked

  • current_annotations (list) – Current annotations (in SVG path format)

save_label(label_name, label_text, image_bbox, slide_information)

Save new label to current save folder.

Parameters:

  • label_name (str) – Name of label type adding label_text to (e.g. “color”)

  • label_text (str) – Label to add (e.g. “blue”)

save_mask(annotations, class_options, image_bbox, slide_information)

Saving annotation mask with annotated classes using pre-specified format

Parameters:

  • annotations (list) – List of SVG paths from the current figure

  • class_options (list) – List of colors and names for all classes

  • image_bbox (list) – Bounding box for the current image

update_slide(slide_selection, vis_data)
update_structure(structure_drop_value, prev_click, next_click, current_structure_data, current_class_value, slide_information)

Updating the current structure figure based on selections

Parameters:

  • structure_drop_value (list) – Structure name selected from the structure dropdown menu

  • prev_click (list) – Previous button clicked

  • next_click (list) – Next button clicked

  • current_structure_data (list) – Current structure bounding boxes and indices

  • current_class_value (list) – Current class value from the class dropdown menu

Returns:

Updated figure containing new structure, updated structure index if previous or next button is clicked, cleared label text

Return type:

tuple

update_structure_options(slide_bounds, current_features, active_tab)

Updating the structure options based on updated slide bounds

Parameters:

  • slide_bounds (list) – Current slide bounds

  • current_features (list) – Current set of GeoJSON FeatureCollections

  • active_tab (list) – Active tab in vis-tools

Returns:

Updated structure dropdown options and bounding box data

Return type:

tuple

Designing Custom Components