Fluorescent Images

Introduction

Fluorescence experiments are almost always captured on highly sensitive monochrome cameras in 12, 14 or 16bit. This means that each fluorophore is captured as a different image with no associated colour and usually cannot be opened in the standard images viewers built into Windows, OS X or Linux. Some default viewers will display 16bit images if they are using the full range of intensities but will not assign any wavelength specific or custom look up tables (LUTs). 

To make the data more presentable it is usually a good idea to add colour to each channel (usually in the colour of the fluorophores emission) and to create composite images of 2 or more of the channels combined.

When merging fluorescent images, you need to be aware that they could appear very strange to someone with colour blindness. Fiji can simulate what your image will look like to someone with colour blindness. 

This method uses the images found in Demo Images\Widefield Images\4 Chanel Fluorescence, Demo Images\Widefield Images\DIC and Fluorescent folders and Demo Images\Colour Blindness folders

Changing Bit Depth – 24bit RGB

Most standard image viewers and presentation software will only show images that are in a 24bit RGB format (3 x 8bit colour channels). Changing a single 12, 14 or 16bit image to 24bit RGB is very simple

1. Open the FITC.tif image found in Demo Images\Widefield Images\DIC and Fluorescent folders. This is a single channel 16bit monochrome image. Earlier versions of windows will open it as a black square. It is best to convert it to a 24bit RGB image for any presentation or publication situations. DO NOT alter bit and colour depth if you plan to do any analysis on the image, always save the results of the conversion as a separate file to preserve your original image.

2. Go to Image 🡪 Type 🡪 RGB Colour 

3. The image, while looking the same, is now a 24bit RGB format that is compatible with pretty much any software out there. Save the file as a copy to preserve your original data.

Adding Colour – Using Lookup Tables

The colour of a monochrome image can be changed easily by changing the lookup table (LUT) applied to it. A monochrome image has a grey look up table applied to it, meaning that each intensity is represented by a different shade of grey. This can be changed to be different shades of red, green, blue etc. LUTs that contain more than one colour can also be useful for showing things like under and over saturation or to better represent signal distribution.

Single Colour LUTs

1. Open the FITC.tif image from before and go to Image 🡪 Lookup Tables. You will see a large list of possible LUTs to apply. Apply the Green LUT. Alternatively, you can press the LUT button on the toolbar.

2. To save a version of the image for presentation etc. follow the steps above to convert the image to a 24bit RGB

Multicolour LUTs

Saturation Indicator

1. Open the FITC (Mod).tif image. Go to Image 🡪 Lookup Tables and apply the HiLo LUT

2. This LUT gives a monochrome image with some pixels in red and some pixels in blue. The red and blue pixels represent intensity values at the two extremes of the range. Red pixels are fully saturated (i.e. pure white) and blue pixels are fully under saturated (i.e. totally black). Both these types of pixels essentially contain no information as you cannot be sure how far above or below saturation they are, they are essentially a cartoon.

Intensity Distribution

1. Open the FITC (Mod).tif image. Go to Image 🡪 Lookup Tables and apply the Rainbow RGB LUT

2. This LUT colours intensity in shades of red, blue or green. The highest 33% of intensities are reds, the middle 33% are greens and the bottom 33% are blues. This gives the viewer the ability to easily see the range of intensities present in the image.

3. A calibration bar of what intensities the colours cover can be added to the image by going to Image 🡪 Tools 🡪 Calibration Bar

4. In the Calibration Bar dialog you can configure the position and configuration of the calibration bar. The Overlay tick box allows you to create the bar as an overlay of the original image, instead of burning it into the image permanently 

Merging Images

Merging or overlaying images of several different fluorophores can provide a very pretty picture that gives the viewer information about relative location, intensity etc. Depending on how many channels, and what colour you want them, are being merged the method is slightly different.

Merging Up To 7 Channels and DIC/Phase/Brightfield – Standard Colours

Fiji/ImageJ has a tool for easily merging up to six channels plus one brightfield/grey channel. This tool will however only allow you to colour your image channels red, green, blue, yellow, cyan and magenta (plus grey for the brightfield). To be able to create a merged image with “non-standard” colours (i.e. not red, green and blue) see the second method below.

1. Open the DAPI, FITC and DIC files from the Demo Images\Widefield Images\DIC and Fluorescence folder.

2. Go to Image 🡪 Colour 🡪 Merge Channels 

3. Assign an image to each channel. In this example the red channel is set to none as there is no red image, FITC is assigned the green channel, DAPI the blue and DIC the grey channel. Un-tick the create composite box and tick the keep source images box. Press OK 

4. The resulting image will be a standard 24bit RGB colour image showing the merge of the channels. 

Merging Up To 7 Channels and DIC/Phase/Brightfield – Non-standard Colours

There will be some instances where you may want to merge the images as above but do not want to have them coloured the standard red, green, blue, yellow, magenta, cyan and grey combination (for example if you are presenting them to a person who is colour blind or you want one of your channels to have a multi-colour LUT applied).

1. Follow steps 1-3 in the method above, but this time make sure the create composite box is ticked. 

2. The resulting image will look the same on the screen as the previous merge image did but there are some differences. 

3.  The image is not a 24 bit RGB image as before, it is now 3 16 bit images on top of each other. There is a slider at the bottom that lets you select the active channel. You tell which channel is selected by the colour of the border and text at the top right of the image.

4.  To change the colour of the one of the channels go to Image 🡪 Colour 🡪 Channels Tool 

5. The resulting dialog box has several options. In its default state it will say composite in the box at the top and have all the channels ticked. Unticking the boxes next to each channel will turn that channel off. The example below shows the DIC channel (Channel 3) disabled.

NOTE: A recent update (January 2022) to ImageJ added 4 choices of composite mode compared to the original 1. Composite Sum is the original mode, the other modes may give better looking images but will be performing some modifications to achieve it. 

6. To change to colour of a given channel use the slider at the bottom of the image to select the desired channel (in this example channel 2 – blue). 

7. Click the LUT button on the main Fiji window and select the colour you would like to apply (Cyan and in this example) 

8. Once you are happy with the image you can convert it to RGB colour by going to Image 🡪 Type 🡪 RGB Colour

7 Colour Merge – Just Enough Channels

When merging 7 colours there is just enough standard colours to show a reasonably easy to understand overlay. The three primary colours of Red, Green and Blue can be used. The three secondary colours of Cyan, Magenta and Yellow can then be sued. Grey is the last choice. After these the choices are either gradients or tertiary colours that fall between the secondary colours like orange, teal etc.

NOTE: There is no combination of 7 colours hat can be made colour blind compatible. It is always recommended to show complex merges as individual channels as well in any publication.

1. Open all 7 images found in the Demo Images\Confocal\7 Channel folder 

2. Go to Image 🡪 Colour 🡪 Merge Channels and assign each image to a channel 

3. The resulting image will be displayed in the composite sum mode which can look odd for large numbers of colours of channels with coincident signal. 

Changing to Composite Max will give a better looking image in this case 

Merging 8 or more colour channels

To create a merge of 8 or more colours a different method has to be used. The images that you want to colour need to be added into a stack.

1. Open the 4 images in the Demo Images\Widefield Images\4 Channel Fluorescence folder. For this example we will only use 4 images, but this method can be used for an essentially infinite number of images

2. Go to Image 🡪 Stack 🡪 Images to Stack 

3. In the resulting dialog box make sure both boxes are ticked. You can give the stack a name, or just leave it as the default “Stack”. The result will be a stack of 4 images, one plane for each channel. 

NOTE: The images will be added into the stack in alphabetical order based on their names (e.g Alexa 488 first, Hoechst last)

4. To be able to change to colour of each plane individually the stack needs to be converted to a composite image. Go to Image 🡪 Colour 🡪 Make Composite. Leave the display mode as composite and press OK in the resulting dialog box. 

5. You now have a composite stack like before that you can assign new LUTs to 

Simulating Colour Blindness

It is possible in Fiji to show what a merged image will look like to someone with varying sorts of colour blindness. The current plugin allows you to simulate Protanopia (no red), Deutanopia (no green) and Tritanopia (no blue). More advanced plugins are available that allow simulation of more subtle forms.

1. Open Colour Blindness.tif from the Demo Images\Colour Blindness folder. This image contains cells stained with DAPI in blue, green with CellTracker and red for a labelled ligand. To some colour-blind people this would not be distinguishable. For example the red and green would look a similar colour with some variation in intensity.

2. Go to Image 🡪 Colour 🡪 Simulate Colour Blindness 

3. Select the type of colour blindness you wish to simulate from the list and press OK. 

4. The results will show what your image may look like to someone with that type of colour blindness. In the example below the middle image is no green and the right hand image is no blue.