New to X-ray μCT?

• X-ray microcomputed tomography (μCT) or microfocus computed tomography is conceptually equivalent to medical CT, where hardware characteristics and arrangements are optimised for high spatial resolution (in the order of 1 to 100 μm), typically used for imaging material and tissue samples ex vivo and in situ, with common sample dimensions in the order of millimeters to centimeters. 

In keeping with medical CT, μCT imaging is accomplished by placing the sample in the X-ray beam path and capturing projected X-ray absorption patterns (radiographs) over a large number of different rotation angles (typically hundreds to thousands). Contrary to medical CT, where the X-ray source and the detector rotate in a gantry system around the patient, in μCT systems, the X-ray source and detector are usually fixed in place and it is the sample that is rotated during image acquisition. 

On completion of a scan, CT reconstruction algorithms are used to derive the X-ray absorption of the sample. The technique was initially developed and optimised to image mineralised bone structures at a microscopic level, and since then μCT is used routinely in many fields, including archaeology, biomedical research, engineering, materials science, and paleontology. 

source: https://doi.org/10.1016/j.ajpath.2019.05.004

A brief introduction to the X-ray microtomography can be found here:
Landis, Eric N., and Denis T. Keane. "X-ray microtomography." Materials characterization 61.12 (2010): 1305-1316.

New to 3D X-ray Histology?

• In the biomedical field, μCT has been successfully used over the past two decades to image biological tissues ex vivo. Tissue imaging applications have also been reported, but these mostly rely on laborious and intrusive sample preparation protocols that entail the use of X-ray attenuating stains (eg, osmium tetroxide or iodine), complex X-ray optics systems, and/or synchrotron light sources.

3D X-ray histology (XRH) is a μCT-based imaging technique that allows non-destructive 3D (volume) visualisation of standard formalin-fixed, paraffin-embedded (FFPE) biopsy specimens and can be seamlessly integrated into conventional histology workflows, enabling nondestructive three-dimensional (3D) X-ray histology.

source: https://doi.org/10.1016/j.ajpath.2019.05.004

For more details please visit our main 3D X-ray Histology site to read through a short introduction about XRH, our journey so far and the road ahead.

Frequently asked questions (FAQ) and answers

• What kind of images are generated during μCT XRH data acquisition?
  A set of 2-D images (usually called angular projections or radiographs) is captured at different angular positions as the sample is rotated during the scan. These projection images contain information about the attenuated X-ray beam, which passes through the sample.
  Grey-scale values of pixels and their variation in angular projection images represent changes in the attenuation coefficient and therefore local density of the sample.

• What kind of 3D datasets are reconstructed from the set of 2D radiographs?
  An acquired set of 2D angular projections is reconstructed into a 3D rectangular grid composed of cuboidal building blocks known as voxels, whose grey-scale values correspond to the linear attenuation coefficients of the sample material in these elementary volumes.
  This 3-D volumetric map of linear attenuation coefficients can also be reconstructed as a stack of 2D cross-sectional slices. CT reconstruction can be used for visualisation and quantification work, e.g. detection, extraction and measurement of features of interest. Volumetric datasets can also be converted into high-fidelity finite element meshes and CAD models.

• What is a pixel?
  A pixel, or picture element, is an element of a two-dimensional regular grid containing a value. Typical picture files might contain 24 bits of information per pixel: 8 bits for each red, green and blue component of that element. Grey-scale images, such as radiographs or sinograms, will store a single value for their representative component. This might be 8, 16, or 32 bits in size.
  The size of the grid is usually given as x x y pixels. By multiplying the number of bits per pixel b with the x and y dimensions, we can calculate the storage consumed by this image (the factor of 1/8 converts bits to bytes).
  n (bytes) = x * y * b/8
  also see https://en.wikipedia.org/wiki/Pixel

• What is a voxel?
  A voxel, or volume element, is an element of a three-dimensional grid: a 3D analogue of the pixel.
  Typically these will contain only one data value each. As with a greyscale image, this might be 8, 16 or 32 bits.
  A grid of voxels is measured as x x y x z voxels. Similarly to pixels, multiplying the number of bits per voxel with the x, y and z components yields the storage consumed by the volume set.
  n (bytes) = x * y * z * b/8
  N.B. the value of a voxel resulting from reconstructed CT data has densiometric significance.
  also see https://en.wikipedia.org/wiki/Voxel

• What kind of data should I expect to be generated by an XRH session?
  A typical XRH dataset of a single sample consists of: 

  1. the raw reconstructed volume data (i.e. the "3D volume" file)

  2. refined raw volume(s); These are processed raw volume file saved in .raw and/or .tiff format, which are resliced to a histology-relevant orientation and/or have been enhanced using noise reduction / contrast improvement and/or ct-artefact removal techniques. 

  3. a number of automatically generated videos that provide a quick overview of the dataset

  4. [optional] other processed volume files saved in .raw and/or .tiff format, which might include segmentation, labelling, etc

  5. [optional] the raw projection data underlying the CT scan

• What is the content of my data folder?
  Your data will be normally organised by their XRH sample ID, and each "sample" folder might contain multiple scans of that sample. The Folder structure will resemble the following:
  --
  [XRH-ID]; e.g. DEMO02019-FFPE
  |-- [Scan name]; e.g. 20210129_XRH_1234_OLK_DEMO02019-FFPE_notes *
  |-- Raw radiographs (.tif) + scan/info files (.xml / .xtekct)
  |-- [Centre slice] **
  |-- [Reconstruction folder]; in a standardised format *
  |-- [videos]; (when applicable) containing videos of the reconstructed volume(s)
  |-- [extra]; (when applicable) containing screen-shots, exports and/or processed data other than refined raw
  |-- Raw reconstruction file (.vol) + Raw reconstruction and scan settings file (.vgi) + Refined raw volume files (.raw / .tif) ***
  
  * in a standardised format YYYYMMDD_ScanerID_JobID_OperatorInitials_SampleID_note-1_note-2_
  ** Automatically generated folder containing a single sinogram and info-files
  *** see relevant section in Fiji - ImageJ getting started for how to handle these files

• Can my data be expressed in Hounsfield Units (HU)?
  In principal HU are not applicable to X-ray μCT images of FFPE samples. HU base unit is the attenuation-of-water, which is set to 0 (air to -1000) to allow for a dissent range and consistency of both soft & hard tissues grey levels [see HU in radiopedia.org]. FFPE samples are dehydrated, so if you express your scans in HU all values will turn negative, which is going to make things complicated with quantification, rendering, etc.. One can "calibrate" their data for consistency using phantoms containing materials of known attenuation or "normalise" them expressing the grey values in relative terms; i.e. all grey-levels of same materials to appear in the same bracket (i.e. Wax = + XYZ, Air = 0, Plastic cassette = +ZXY). Calibration is a complex task. It goes into the XCT metrology territory and it get very complicated very quickly. Normalisation is simpler, especially if you are willing to accept some basic approximations. One for example can use the grey-value of Wax, Air and Plastic to generate a normalisation curve and normalise their data by assigning specific values to these "entities". 

All datasets generated at the XRH facility are managed using an in-house designed management system called ‘XRH management system’ (XRHMS). Amongst other functionalities, XRHMS can automatically extract metadata from the different datasets, present summary information (such as imaging parameters) and representative images. By doing this, user-friendly PDF reports can be automatically generated and presented to the end user. 

The system allows two levels of reporting:

• Level 1 - Automatic reports generated upon scan completion, which include the imaging parameters and a representative single central slice through the automatically reconstructed volume; such a report can be seen here: Example of Level 1 Report example

• Level 2 - This is an operator-triggered process, featuring videos of the data set, which can include single and/or "thick" -slice roll-throughs of the sample in any plane, 3D renderings for the µCT data sets, etc. Upon generation, the videos are automatically published online and are assigned a unique url. These provide an immediate overview of the data for biomedical and clinical researchers and their collaborators.
  see: Example of Level 2 Report example 

For a quick guide on how to read a PDF report, please see the 'PDF report explained' file below:

PDF_report_explained_generic_v3 - by Dr. Elena Konstantinopoulou and the XRH team

The imaging equipment and computing resources available at the X-ray Imaging Centre and at the Biomedical Imaging Unit at the University of Southampton provide a vast array of imaging, visualisation, analysis and simulation opportunities.

Both µ-VIS and BIU offer consultation and training sessions which aim to enable our users to make the best of these opportunities.

• Training at μ-VIS 

• Training at BIU

If you are interested in attending any of these please get in contact with us.

For "getting started" material and guidance on specific software that is relevant to XRH visualisation and/or analysis, please see dedicated Software page.

For guidance on appropriate interpretation of the XRH images and information about the different viewing modes you could come across after receiving your XRH datasets please see dedicated Viewing modes page.

It is highly recommended that you read this section alongside Artefacts section.

please see relevant page here