How to get Started

In this video, I’m going to show you how to rapidly get started using MicrobeJ

You’ll find the MicrobeJ Plug-in is located in the ImageJ Plugins menu

To orient you to the layout of MicrobeJ,

First, You’ll notice that Microbe J has several tabs along the top which allow you to set up an experiment

load your image,

define parameters for cell detection and analysis,

detect fluorescent foci associated with the bacterial cells,

use templates which speed up analyzing image sets and

and finally track any errors that might arise during analysis.

In this video, we will explore the features in the Experiment, Images and Bacteria tabs by starting with this stack of phase contrast images and show you how to optimize detecting cells, extract simple information like length and width, perform simple statistical analysis and graphically represent your data.

On the experiment tab, you can specify a name and a description of your experiment, but it's not mandatory.

An experiment is just an image or set of images that you want to analyze with specific settings.

We currently have a stack of phase contrast images open on the desktop

To select this stack of images in MicrobeJ, go to the Image tab and select “image” as the input mode.

There are a lot of other ways to open or select images in MicrobeJ, but we’ll explore those later.

Now Go to the ‘Bacteria’ tab.

In this panel, we will select the channel of interest, in our example we only have 1 channel

Next, We will select the type of background, In this case, the background is brighter than the particles of interest so let’s keep the ‘Bright’ mode.

Finally, we will select the automatic thresholding method.

Most of the time you can select ‘Default’ as the automatic thresholding method, but there are times when you might need to select a more appropriate method.

For example, if the background is unevenly illuminated you might want to use local thresholding methods.

If you are interested in any of the other thresholding methods listed in the drop down menu, please take a look at the documentation available on our website; you’ll find everything you need there to properly use those methods.

Note that you can display or hide the thresholded values on the image by clicking on this button.

You can also set an offset on the threshold level by changing the position of this slider or using the scroll wheel on your mouse when the cursor is positioned over the slider. Watch as I move the slider,  how the thresholded values in yellow change on the image. Another way to set an offset by putting a numerical value in this box.

Now we will move down to the morphology detection panel and select a mode of detection. You can see that there are several choices, but for this example we will use the ‘Medial Axis’ mode and click on the test button.

When you click on the test button, the particles will be detected on the active slice of your image stack using the current settings.

If we test our settings on slice 5 of our image stack, we will notice that not every particle detected will be a bacterial cell.

So you might ask, how do I know what settings to change to optimize detecting only the particles of interest that are cells?

If you zoom in on your image, you’ll see that,

for every detected particle, you have access to their specific attributes:

such as their area, width, or length.

You can use this information to help you easily define what particles to include in your analysis.

So if you want to exclude this tiny particle for instance from the final analysis, just adjust the minimal area to something greater than 0.4, and less than 1 click on the test button again.

As you can see, the particle has turned red and this means that it is now rejected from the final analysis.  

Using this same procedure, you can refine the rejection process by using other attributes such as the width or length but also more complex attributes such as the angularity or sinuosity

If you want more details about those different attributes, please take a look at the documentation available on our website.

Note that you can customize the colors used for displaying the contour of the detected cells, or the rejected particles as well as a lot of other features, such as the medial axis, poles or sides by clicking here the color pallet button.

When you are done optimizing cell detection, you can now select different options to analyze your cells.

For example if you are interested by the length, the width or any shape attribute select ‘shape descriptors’ and all those attributes will be listed in the raw data table.

Another option that is very useful is “segmentation” which can be used to separate cells that are touching each other, let’s select that option now and go to the first slice in our stack of images and press test, You can now see that the two cells that were initially rejected from our analysis have now been segmented and will be included in our final analysis.

MicrobeJ offers a lot of other very interesting options, such as septum detection to detect septa along the medial axis of the cell, or profile, when you want to extract the intensity profile along the medial axis of a cell or to simply get the image of the cell…. but we’ll talk about those options later.

Up until now, we have only tested our settings on a single slice of our stack of images, but I’d like to point out that some of the buttons in MicrobeJ have a dual role that is indicated by this shaded lower right corner on the button.

This means you can access an additional related function through the same button.

On the test button, for instance, if I use the scroll wheel on my mouse, or 2 finger scroll feature on a trackpad, you can change the test button to test all the slices in the stack, not just the active slice. Now when we move through the stack of images, we see that all cells have been detected.

If you will be analyzing images of similar types of bacteria in the future, it might be useful to reuse these settings. To do that, you want to save these settings into a file.

At any time you can save these settings as an xml file by clicking on the ‘save as‘ button. You can also load existing settings by clicking on the ‘open’ button and select your settings file or simply drag and drop the file on the MicrobeJ status bar. Note that you can also drag and drop images, templates, experiments and results files here.

Ok, now that we have defined and tested our settings, and saved them to use for a future image analysis, it’s time to run the analysis on the stack of images and obtain the raw data.

When you click here you will be presented with the MicrobeJ’s results interface.

The results interface combines all the information about the experiment and the bacteria that MicrobeJ has detected.

To get access to any information associated with the bacteria, just click on bacteria here on the left panel and you will see a list of all the bacteria detected in that image on the right panel.

Each row is an individual bacterium detected in the image and is directly linked to the image. So if you click on one row, it will highlight the corresponding bacteria detected on the image and likewise if you select a bacterium on the image, it will highlight the corresponding row of the list. This makes it is easy to interact with your data set to temporarily exclude, include or completely remove bacteria from the list and ultimately from the final analysis.

By default, to simplify the display you only see a limited number of columns in the list, but each of these columns contains a collection of features associated with the column name. For instance, in our example, we checked the shape descriptor option when we set up the experiment, you will see a column called Shape. By right mouse button clicking or holding down the control button and clicking on the column title, this displays a submenu with the available features such as area, length, width and other shape attributes. By clicking on one of these, it adds a new column to the table with the corresponding feature; however, you can also hide a column in the same manner, by right clicking on the column and use the HIDE COLUMN function.

Once you’ve revealed the data that you want to analyze further, you have a several options.

At anytime, you can right mouse click or control click anywhere on the list and copy the active data set to the clipboard to paste it into your favorite statistics or graphing software tool or save it as a file.

But, I’ll have to admit that one of the most powerful features of MicrobeJ is that you don’t have to leave the MicrobeJ working environment to run statistical analyses or graphically represent your data.

MicrobeJ has you covered! This functionality is built in!

In this getting started with MicrobeJ video, I’m only going to give you a glimpse of what the results interface has to offer because it can do so many things that you would normally require several dedicated software packages to achieve that it really deserves a video of its own to explore them in more depth.

So let’s just do a few simple things like calculate the mean of the cell lengths for all the bacteria you detected in your stack of images

and then plot a histogram of the distributions of the lengths.  

In the process we can also learn how to save analysis and graphing templates that can be loaded for future image analysis sessions.

You will see that the Results interface is very easy to use and highly customizable.

To orient you to the Results interface, let’s look at the layout.

In this panel, it shows the Experiment and all the detected bacteria; this panel would also show if the bacteria had fluorescent foci or filaments associated with them.

In the panel below this, any templates you have used or generated will show here.

The buttons at the top left provide you with options to add existing results to a previous experiment or save your results. The buttons above the main raw data panel, allow you to see your raw data, access the statistics or chart generators, perform different types of regression analyses and access data representation modes.

Below these buttons are boxes to allow users to add columns to the data table, which may contain custom calculations

Finally, the buttons on the top right, allow you to add a column, refresh columns and pop windows out containing different subsets of highlighted data, statistics or charts.

In our example, we will only use a few of these options.

To get the average length of your bacteria in the population, you can go to the statistics button,

select mean in the statistics dropdown menu and

select the feature that you want to display,

in our expt, this would be the Shape attribute Length and press PLAY and you get access to that value directly in the table.  You can pop the statistics result out into a separate window using the pop out to window button.

Likewise, if you are interested by a graphical representation of your raw data you can go to the charts button located above the raw data table, select histogram, select the shape.length column and press PLAY.

If you want to save this chart, click on the pop out to window button.  From here you can change the labels, title, color scheme, scale, and resize the chart by right mouse clicking on the chart, selecting properties from here, then once you are happy with the way it looks, right mouse click on the final chart and save it.

If you want to try graphing your results different ways and altering the appearance without losing each one, you can generate different charts de novo in the chart generator.

When you get the one you like and would want to use again in the future, you can save this as a template. Each of the pop out windows becomes a template that can be saved and are listed on the results interface at the bottom left panel.

So in this video, you should have learned how easily it is to use MicrobeJ to load images, optimize the detection of bacteria in those images, get your raw data and perform analyses and graph them.  

Thanks for your interest in the MicrobeJ project!

Stay tuned for more!

Created with the Personal Edition of HelpNDoc: Free EPub and documentation generator