Aivia Software

Pixel Colocalization

Owned by Hoyin Lai
Last updated: Jan 06, 2023 by Aiden Maloney-Bertelli (Unlicensed)Version comment
6 min read

The Pixel Colocalization recipe is an image enhancement function that detects fluorescent labels from two input channels of the same image that are spatially overlapped (or "colocalized"). The recipe extracts pixels above the user-defined threshold while suppressing the rest. The result output can be used for additional processing using other Aivia detection recipes.

By default, the recipe generates two output channels: Colocalized Pixels Ch1 and Colocalized Pixels Ch2 containing the pixels that are above the specified intensity threshold from the respective input channels. The user has the option of adding a third output channel, Combined Coloc Channel, that includes only pixels that are spatially overlapped between the two inputs, with the pixel intensity the average of the two input channels.

 

Inputs and Outputs

The Pixel Colocalization recipe takes in three input channels and can generate up to three output channels. An output channel can be turned off by de-selecting the checkbox next to its name in the Input and Output section.

Inputs

There are three inputs to the recipe, and their descriptions are summarized in the table below.

Input Channel

Description

Input Mask Channel

This input defines the image region that the colocalization analysis will be applied to. The Mask channel serves as an initial region-based filter for colocalization (for example: to demarcate cells from extracellular regions when analyzing spatial overlap between intracellular proteins). Colocalized signal that falls outside of the Mask channel region will not be included in the recipe output.

Typically, this input is the same as one of the two input colocalization channels.

Input Coloc Channel 1

This input specifies the first of the two input image channels used for analyzing pixel colocalization.

Input Coloc Channel 2

This input specifies the second of the two input image channels used for analyzing pixel colocalization.

An example of how each input channel is used is shown on the image on the right.



Input examples for Pixel Colocalization. The Input Mask Channel can be used for identifying the area the colocalization is applied, e.g. the whole cell (green). Input Coloc Channels 1 and 2 are used for colocalization, e.g. identifying spatial overlap between mitochondria (red) and the endoplasmic reticulum (blue). Red and blue pixels that are colocalized outside of the cell boundaries will be discarded. Image Courtesy: Broad Bioimage Benchmark Collection.

 

Outputs

There are three possible outputs from the recipe which can be specified in the Input and Output section of the recipe. Descriptions of the outputs are in the table below.

Output Channel

Description

Output Channel

Description

Colocalized Pixels Ch1

This output contains all image regions from Input Coloc Channel 1 that are above the user-defined threshold; pixels that are below the threshold or outside of the defined Mask Channel area are suppressed.

This output is turned on by default.

Colocalized Pixels Ch2

This output contains all image regions from Input Coloc Channel 2 that are above the user-defined threshold; pixels that are below the threshold or outside of the defined Mask Channel area are suppressed.

This output is turned on by default.

Combined Coloc Pixels

This output contains all image regions whose pixel intensities are above the user-defined threshold for both Input Coloc Channels 1 and 2 (or colocalized). The pixel intensity of this output is the average intensity of both Input Coloc Channels.

 

Parameters and Presets

Parameters

Recipe parameters for Pixel Colocalization and their descriptions are summarized in the table below.

Preset Group

Parameter Name

Min Value

Max Value

Description

Preset Group

Parameter Name

Min Value

Max Value

Description

Mask Detection

Min Intensity Threshold

0

255 (8-bit)

65,535 (16-bit)

Specifies the minimum intensity for image region detection; the image region defines the area that the colocalization analysis will take place; a lower value will incorporate a larger area of the image for colocalization analysis

Fill Holes Size

0

50,000

Adjusts the maximum size threshold for filling in gaps inside the detected image region; a lower value will preserve more holes in the detection

Smoothing Factor

1

100

Adjusts the amount of smoothing applied to the surface reconstructions of the detected image region; a lower value will preserve more of the region's morphological features

Channel 1 Detection

Min Intensity Threshold

0

255 (8-bit)

65,535 (16-bit)

Specifies the minimum signal intensity for the first of the two input image channels used for colocalization analysis; a lower value will detect more, dimmer signals

Min Colocalized Pixels

0

50,000

Specifies the minimum area of contiguous pixels to be included in the analysis output

Channel 2 Detection

Min Intensity Threshold

0

255 (8-bit)

65,535 (16-bit)

Specifies the minimum signal intensity for the second of the two input image channels used for colocalization analysis; a lower value will detect more, dimmer signals

Min Colocalized Pixels

0

50,000

Specifies the minimum area of contiguous pixels to be included in the analysis output

 

Presets

There are three preset groups in the recipe: Mask Detection, Channel 1 Detection and Channel 2 Detection; each group has three pre-configured parameter groupings to help you get started on the analysis. The default preset values are as follows:

 

Mask Detection

Parameter Name

Low

Medium

High

Intensity Threshold

64 (8-bit)

20 (8-bit)

5 (8-bit)

16,384 (16-bit)

5,243 (16-bit)

1,311 (16-bit)

Fill Holes Size

10

10

10

Smoothing Factor

1

2

2



 

Channel 1 Detection

Parameter Name

Low

Medium

High

Min Intensity Threshold

64 (8-bit)

20 (8-bit)

5 (8-bit)

16,384 (16-bit)

5,243 (16-bit)

1,311 (16-bit)

Min Colocalized Pixels

1

5

10



 

Channel 2 Detection

Parameter Name

Low

Medium

High

Min Intensity Threshold

64 (8-bit)

20 (8-bit)

5 (8-bit)

16,384 (16-bit)

5,243 (16-bit)

1,311 (16-bit)

Min Colocalized Pixels

1

5

10



 

Tutorial

Before beginning the tutorial, please download the Pixel Colocalization Demo Image. For information on how to select presets or modify parameter values, please refer to the tutorial on how to use the Recipe Console.

  1. Unzip the demo file and load the demo image, PixelColocDemo.aivia.tif, into Aivia

  2. Click on the Image Enhancement Tools tab on the side panel to the right of the image. Click on the Recipe selection dropdown menu and select the Pixel Colocalization recipe

  3. Click on the caret titled Input and Output to select the input and output channels. Use the following channel settings:

    • Input Mask Channel: Ch3-T3

    • Input Coloc Channel 1: Ch2-T2

    • Input Coloc Channel 2: Ch3-T3

  4. Select the Medium preset for the Mask Detection preset group; and the Low presets for both the Channel 1 Detection and Channel 2 Detection preset groups

  5. Click on the Show Advanced Interface icon  to expand the recipe settings

  6. Change the parameters listed below to the specified values, leaving the others intact:

    • Mask Detection

      • Min Intensity Threshold: 800

      • Smoothing Factor: 6

    • Channel 2 Detection

      • Min Intensity Threshold: 1500

  7. Click the Apply to All Frames button to begin applying the recipe to the image

Two output channels are created by default, consisting of the output to Input Coloc Channels 1 and 2 respectively. You can add the Combined Coloc Channel output to the image by clicking on its checkbox in the Input and Output section.

 

Results

 

Animated GIF showing the three individual Pixel Colocalization outputs (Colocalized Pixels 1, Colocalizeds Pixel 2, Combined Coloc Channel), combined Colocalized Pixels 1 and 2, and the input image



 

Image Courtesy

Image set BBBC022v1 [Gustafsdottir et al., PLOS ONE, 2013], available from the Broad Bioimage Benchmark Collection [Ljosa et al., Nature Methods, 2012].

Kasturi Mitra, Cell Image Library (CIL:9069), http://cellimagelibrary.org/images/9069



 

Related articles