Fig 1.
Geometry of the micro-CT system.
(a) Top view depicting the arrangement of the internal components of the system: x-ray source, rotary stage, phosphor detector screen, detector-screen tilt angle (DSTA), lead plate, camera, lens, source-to-object distance (SOD), and object-to-detector distance (ODD). The orange dotted-lines represent the trajectory of the polychromatic, cone-shaped, x-ray beam. (b) Perspective view of the system showing the arrangement of the various 20 x 20 mm aluminum extrusion profiles. One of the opaque panels was removed from the 3D rendering to allow visualization of the internal components.
Fig 2.
Experimental setup using the general-purpose x-ray source.
Note that the rotary stage is not visible as it is hidden behind the lead plate. The source-to-detector distance (SDD) was 600 mm.
Table 1.
X-ray source comparisons and exposure-level matching, x-ray protocols.
Table 2.
Summary of the estimated costs for the main components of the low-cost CT scanner–June 2021.
Fig 3.
Summary of image processing steps required between projection data acquisition and CT reconstruction.
Fig 4.
Assessment of geometric distortion correction resulting from the tilted-detector configuration.
(a) shows the un-corrected raw data of the Cartesian calibration grid. (b) corrected image of the Cartesian calibration grid.
Fig 5.
Physical appearance of the imaged phantoms for this study.
(a) comprehensive quality assurance CT phantom, (b) porous, gyroid-based, cylindrical, titanium-alloy (Ti6Al4V) phantom, and (c) titanium-alloy (Ti6Al4V) resolution phantom.
Fig 6.
Computer-aided design (CAD) of the porous gyroid-based cylindrical titanium-alloy (Ti6Al4V) scaffold.
(a) perspective view of 3D rendering of the CAD. (b) Trans-coronal synthetic slice of the porous cylinder CAD. (c) Trans-sagittal synthetic slice of the porous cylinder CAD. (d) Transaxial synthetic slice of the porous cylinder CAD. (b),(c), and (d) are color coded to depict the titanium-alloy (yellow) and internal porosity (red) regions of the scaffold.
Fig 7.
Computer-aided design (CAD) of the resolution titanium-alloy (Ti6Al4V) phantom.
(a) translucent perspective view of 3D rendering of the phantom’s CAD. (b) Trans-coronal synthetic slice of the phantom’s CAD. (c) Trans-sagittal synthetic slice of the phantom’s CAD depicting details of the internal voids. The bar patterns are described in lp mm-1 all the other nominal measurements are shown in mm (d) Transaxial synthetic slice of the phantom’s CAD. Note that the labelled measurements represent the prescribed nominal values; these values may not be achieved during the manufacturing of the part.
Fig 8.
Results of geometric calibration evaluation using a cylindrical phantom with four radiopaque marker beads.
(a) x-ray projection radiograph of the phantom; (b) uncorrected and corrected elliptical trajectories prescribed by the phantom’s markers over the course of one full rotation of the rotary stage. (c) close-up of uncorrected and corrected elliptical trajectories of the top-most marker bead. The corrected position of the markers aligns with the ideal trajectory calculated using a sinusoidal best fit.
Fig 9.
Evaluation of spatial resolution of the cost-efficient CT scanner.
(a) a single reconstructed transaxial-slice CT image of the slanted-edge plate of the CT quality assurance phantom. (b) reconstructed transaxial-slice CT image of the resolution coil plate of the CT quality assurance phantom illustrating the qualitative display of resolution coils.
Fig 10.
Modulation transfer function (MTF) of the cost-effective CT scanner measured from the slanted-edge plate, the resolution coil plate, and the copper wire of the CT quality assurance phantom.
The 10% MTF level was reached at 2.12 mm-1, corresponding to a spatial resolution of 235 μm.
Fig 11.
(a) reconstructed transaxial-slice CT image of the geometric-accuracy plate of the CT quality assurance phantom with four beads located at the periphery and 34.98 ±0.06 mm apart and one central bead at a distance of 24.74 ±0.01 mm from the other four. (b) all possible nominal versus CT calculated distances between beads depicting error bars that represent the 95% confidence interval.
Table 3.
Euclidean distances, in mm, between the centroids of all five beads of the geometric accuracy plate of the CT quality assurance phantom.
Fig 12.
Reconstructed transaxial-slice CT images of the linearity plate (a) with air, water, and various concentrations of iodine shown in mg ml-1; and the CT number evaluation plate (b) of the CT quality assurance phantom.
Fig 13.
Plot of measured CT number within ROIs placed in each iodine vial versus known iodine concentration within the linearity plate of the CT quality assurance phantom.
A significant linear correlation is seen between CT number in HU and iodine concentration in mg ml-1.
Fig 14.
(a) Reconstructed transaxial-slice CT image of the uniformity plate of the CT quality assurance phantom. (b) Radial signal profile taken through the center of the uniformity plate as illustrated by the white dotted-line in (a).
Fig 15.
CT reconstruction of the porous gyroid-based cylindrical titanium-alloy (Ti6Al4V) scaffold using the cost-effective CT scanner.
(a) perspective view of volumetric rendering using the FWHM threshold used to segment titanium. (b) Trans-coronal slice CT reconstruction of the porous cylinder. (c) Signal profile across a wall of the porous cylinder used to determine the FWHM threshold for segmentation. (d) Trans-sagittal slice CT reconstruction of the porous cylinder. (e) Transaxial slice CT reconstruction of the porous cylinder and line profile used for (c). (f) close-up version of (e) illustrating boundaries between the titanium and air ROIs used to measure the porosity of the porous, cylindrical, titanium scaffold.
Fig 16.
Beam-hardening corrected CT reconstruction of the titanium-alloy resolution phantom using the cost-effective CT scanner.
(a) Volumetric rendering of the phantom with the top cap clipped to reveal the internal-void features of the resolution phantom. (b) Transcoronal slice of the resolution phantom showing the relative position of the clipping plane for (a), (c) and (e). (c) Close-up of (e) illustrating the internal features of the phantom at the level of the clipping plane where the outer diameter of the object is 10 mm. Note that the CT-measured true dimensions are labelled in parenthesis after the nominal dimensions of the internal voids. (d) Transsagittal slice of the resolution phantom showing the level where the resolution limit was reached. (e) Transaxial slice of the resolution phantom showing details of the internal features of the phantom at the level of the clipping plane labelled in (b). (f) Transaxial slice at the level of the resolution limit where the external diameter of the phantom was 13 mm. At diameters greater than this, internal-void features are difficult to distinguish and photon starvation artifacts start to dominate.