Extracting Metrics for Three-dimensional Root Systems: Volume and Surface Analysis from In-soil X-ray Computed Tomography Data

The overall goal of this procedure is to extract quantitative information on plant roots from in-soil X-ray computed tomography data. This method of investigating the root structure of live plants can help answer key questions in the biology field, such as how plants respond to changing environmental conditions like drought, microbial activities, and so forth. The main advantage of this technique is that it is non-invasive, therefore, it allows us to study plants in their native environment, the soil.

Visual demonstration of this method is valuable, as the procedure involves numerous steps, including tomographic imaging, data reconstruction, and analysis, which may be obscure without visual aid. To begin, place a potted plant on the sample manipulator of the X-ray tomography scanner at a distance desired for target magnification. For a plant in a two inch diameter holder, set a sample to a source distance of about three inches.

In the instrument control software, set the X-ray power settings to 85 kilovolts and 190 microamps. Then, set the exposure time to one second for a better signal to noise ratio. Set the number of projections to 3, 142 and the number of frames per projection to four for good data statistics.

Using the scan parameters just set, select the Shading Correction”tab. Then click on New”and name the shading correction file. Then click Present”to confirm the new conditions.

Next, click Create”and click OK”to using two reference images. Then click Generate. Then, when the warning popup window appears, physically remove the sample from the beam path before clicking OK.The shading correction will now run.

Now, click on the Acquisition”tab. Name the data file, then select None”for X-ray filter. Ensure that the Minimise ring artefacts”checkbox is checked so that the sample will be rotated in angular steps while the projection images are acquired.

Finally, start the scan by clicking the Acquire”button. After acquiring the raw data, load it into the reconstruction program. Compare the first and last images by clicking on Blink”to make sure the specimen did not move or the scan settings did not change during data acquisition.

Calculate the center of rotation, or COR, by selecting the Center of rotation”tab and using options Automatic”COR finding with High Quality”accuracy and Dual”slice selection for COR calculation. Then, click Start”to run the calculation. After the COR is found, click the Volume”tab to select the sample to be reconstructed.

Use the thumbnails to edit the volume selection windows. Then click Start”to perform reconstruction to create the volume file containing 3D data. To prepare a RooTrak processable image stack, load the volume file into ImageJ by clicking File, Import, Raw, then selecting the volume file.

To optimize the image contrast between the root and soil, click the Image”tab, then choose Adjust Brightness/Contrast”and select Auto. When the region of interest is visible in clear, the settings are considered optimized. Save the images as an image stack in JPEG, BMP, or PNG format by clicking Save As”and choosing Image Sequence.

To process in RooTrak to segment the root, select the Tools”tab and choose Tracker. Then double click on the white entry area of the dialogue box to load an image stack. After the images are loaded, choose the Tracker”tab and set seed points inside the root by clicking several points inside each of the pertinent root sections visible in the top view slice of the volume data.

Set the tracker parameter smoothness to 0.3 and similarity to 0.8. Click the field below Output Directory”to choose a file folder. Then, click Select Folder”and click Start.

The tracking function will then run. This will follow the root from the top image slice all the way to the bottom slice. After opening and viewing the volume data, select the number of slices according to the usable data volume.

In the example shown here, tracking was stopped at 200 slices, equivalent to a depth of 6.2 millimeters, where root boundaries became ill defined. To carry out volume and surface analysis in ImageJ, convert the image stack from RooTrak into a binary image format by selecting Process, Binary, then Make Binary. Use the open source ImageJ plugin, BoneJ, to create the triangular mesh by selecting Plugins, BoneJ, then Isosurface.

In the options window that opens up, set Resampling”and Threshold”to six and 120, respectively. Then, check Show surface”and press the OK”button. In the 3D viewer that opens up, click on the File”tab.

Then click Export surfaces”and save as STL binary. Finally, open IMESHJ. Load the STL file by clicking Select STL File”and enter the voxel size in microns.

Click Enter Voxel Size”and enter 31 microns, then click OK”twice. Click Calculate Surface Area”to acquire the total sample root surface area in square millimeters. Then, click Calculate Volume”to obtain the total sample root volume in cubic millimeters.

The specimen, consisting of two stems of the native grass prairie dropseed, Sporobolus heterolepis, and the original soil around it, is shown here. Voxel size of the reconstructed data generated was approximately 31 cubic microns. After creating a stack of images and increasing the contrast between the roots and soil, the reconstructed data showed that the root and some of the components of the soil have very similar X-ray attenuation factors, resulting in little to no grayscale contrast in the images.

This figure shows a representative seed point that was selected inside a pertinent root section visible in the top slice of the volume data. Using the RooTrak parameters described in this video, RooTrak successfully segmented the selected approximately 200 slices of volume data, which was equivalent to a depth of 6.2 millimeters. Segmentation of the root is also demonstrated in this animation.

Here, ImageJ was used to generate a triangular mesh isosurface of the 3D volume from the data produced by RooTrak. The default settings used produced a detailed isosurface in a relatively quick manner. Finally, using IMESHJ, the surface area was calculated as 351.87 square millimeters, and the calculated volume was 47.27 cubic millimeters.

Once mastered, this procedure can be done in five hours if it is performed properly. While attempting this technique, it is important to remember to follow all safety procedures. It is also important to remember that root segmentation is only possible if there is sufficient X-ray contrast between root and the soil.

Following this procedure, other specimens with branching structures like tree leaves, animal and human lungs, et cetera, can be imaged and analyzed in order to answer additional questions, like how living tissues grow or how they tolerate certain conditions. After its development, this technique paves the way for researchers in the plant science field to explore things like drought resistance in model plants such as Brachypodium. After watching this video, you should have a good understanding of how to image a plant specimen using tomography, how to visualize its root structure and extract quantitative information on the root from in-soil X-ray computed tomography data.

Don’t forget that working with X-ray instrumentation can be hazardous, and the relevant safety procedures should be followed while attempting this technique.