Fluroview

A Python package for visualization of biological data. It is built on top of matplotlib, allowing it to use all its functions. Fluroview is a simple way to generate beautiful and publication-ready figures, especially for multi-channel images representing them as composite images similar to what can be achieved using an Image processing software like Fiji. Alongside generating figures, it also completes the figures with appropriate scalebars, titles and colormaps.

Installation:

You can install this package directly from pip:

pip install FluroView

This will install the latest 0.4.1 version.

Main modules:

• bioviz_multichannel.py: MultiChannelImage is designed to handle multi-channel image data, allowing for easy manipulation and visualization of individual channels or composite images.

  • Key features:

    • Support for single or multiple image channels
    • Flexible colormap assignment, including custom and built-in colormaps
    • Internally intensity rescaling options so that the colors are more visually clear.

Basic Usage:

To create an MultiChannelImage object:

Here we are creating a figure of a 4 channel Rat Hippocampal neuron image (Taken from ImageJ sample library):

eximg = imread('Rat_Hippocampal_Neuron.tif')
exim = MultiChannelImage(
    channels=[eximg[0], eximg[1], eximg[2], eximg[3]],
    channel_names=['Bungarotoxin','ACh','CFP','Hoeschst'],
    colormaps=['pure_red','pure_green','pure_magenta','pure_cyan']
)
#Lots of customization, use or can use panel to plot the figure as well
fig, ax = MultiChannelImage.create_multichannel_figure(
        exim, 
        pixel_size=0.16,
        units='um',
        panel_label='A',  
        rescale=True, 
        label_position='top-right',
        label_color='black',
        channel_label_show=True,
        scalebar_color='white',
        scalebar_font_size=12,
        panel_label_color='white',
        panel_label_font_size=16,
        channel_label_font_size=10,
        scalebar_position='lower right',
        panel_label_position='upper left',
        figsize=(15,5)
    )

Output:

• bioviz_imagepanel.py: The ImagePanel class facilitates the creation of figure panels containing multiple MultiChannelImage objects, ideal for creating publication-ready figures.

  • Key features:

    • Arrange multiple images in a grid layout.
    • Figure layout is done automatically so no need to fidget around with scaling options.
    • Add scale bars, panel labels, and titles.
    • Customize frame visibility and title colors
    • Easy saving of high-resolution figures.

Basic Usage:

Create MultiChannelImage objects for all the images to put together:

exim1 = MultiChannelImage(
    channels=[eximg[0]],
    channel_names=['Bungarotoxin'],
    colormaps=['pure_red'])

exim2 = MultiChannelImage(
    channels=[eximg[1]],
    channel_names=['ACh'],
    colormaps=['pure_green'])

exim3 = MultiChannelImage(
    channels=[eximg[2]],
    channel_names=['CFP'],
    colormaps=['pure_magenta'])

exim4 = MultiChannelImage(
    channels=[eximg[3]],
    channel_names=['Hoeschst'],
    colormaps=['pure_cyan'])

Now use the ImagePanel Class:

panel = ImagePanel(1,5,figsize=(15,5))
panel.add_multichannel_image(0, 0, exim1, title='Bungarotoxin', panel_label='A')
panel.add_multichannel_image(0, 1, exim2, title='ACh', panel_label='B')
panel.add_multichannel_image(0, 2, exim3, title='CFP', panel_label='C')
panel.add_multichannel_image(0, 3, exim4, title='Hoeschst', panel_label='D')
panel.add_multichannel_image(0, 4, exim, title='Overlay', panel_label='E',scalebar=True, pixel_size=0.16,units='um',set_title_color_default=True)
panel.adjust_layout()
panel.show();

Output

The call panel.adjust_layout() automatically does all the necessary scaling and spacing for the given figsize and the panel grid.

Advanced Features

Custom Colormaps

You can use custom colormaps, including hex color codes:

image = MultiChannelImage(
    channels=[channel_data],
    channel_names=['Custom'],
    colormaps=['#FF5733']  # Using a hex color code
)

Grayscale images

The library handles grayscale images and allows applying colormaps for visualization:

grayscale_image = np.random.rand(100, 100)
image = MultiChannelImage(
    channels=[grayscale_image],
    channel_names=['Grayscale'],
    colormaps=['viridis']
)

Panel Customization

You can customize various aspects of the image panel:

panel.add_multichannel_image(
    0, 0, image,
    title='Custom Image',
    panel_label='A', #change label of each panel
    show_frame=False,
    set_title_color_default=True, #set the color of the titles if not it defaults to colormaps of the current image
    scalebar=True, #turn scale bar on/off
    pixel_size=0.05,
    units='μm'
)

To Do

• Release MultiChannelImage
• Release ImagePanel
• Release notebook tutorial
• Add to PyPI
• Release interactive elements with ipympl, scrolling along z-axis, intensity sliders.
• Release simple interactive segmentation using Segment Anything and matplotlib ipympl backend
• More future: add more customization as per needs.

Contributing

Contributions to the FluroView library are much welcomed. Feel free to send pull requests.

License

This project is licensed under the MIT License - see the LICENSE file for details.