Affiliations: [a]
Imaging Research Laboratories, Robarts Research Institute, Schulich School of Medicine & Dentistry, Western University, London, ON, Canada
| [b]
Department of Earth Sciences, Western University, London, ON, Canada
| [c]
Department of Physiology and Pharmacology, Schulich School of Medicine & Dentistry, Western University, London, ON, Canada
| [d]
Department of Surgery, Schulich School of Medicine & Dentistry, Western University, London, ON, Canada
| [e]
Department of Medical Biophysics, Schulich School of Medicine & Dentistry, Western University, London, ON, Canada
Correspondence:
[*]
Corresponding author: D.R. Edey, Department of Geological Sciences, University of Texas at Austin, Austin, TX 78712, USA. Tel.: +1 512 471 9286; E-mail: [email protected].
Abstract: BACKGROUND:X-ray computed tomography (CT) can non-destructively examine objects by producing three-dimensional images of their internal structure. Although the availability of biomedical micro-CT offers the increased access to scanners, CT images of dense objects are susceptible to artifacts particularly due to beam hardening. OBJECTIVE:This study proposes and evaluates a simple semi-empirical correction method for beam hardening and scatter that can be applied to biomedical scanners. METHODS:Novel calibration phantoms of varying diameters were designed and built from aluminum and poly[methyl-methacrylate]. They were imaged using two biomedical micro-CT scanners. Absorbance measurements made through different phantom sections were fit to polynomial and inversely exponential functions and used to determine linearization parameters. Corrections based on the linearization equations were applied to the projection data before reconstruction. RESULTS:Correction for beam hardening was achieved when applying both scanners with the correction methods to all test objects. Among them, applying polynomial correction method based on the aluminum phantom provided the best improvement. Correction of sample data demonstrated a high agreement of percent-volume composition of dense metallic inclusions between using the Bassikounou meteorite from the micro-CT images (13.7%) and previously published results using the petrographic thin sections (14.6% 8% metal and 6.6% troilite). CONCLUSIONS:Semi-empirical linearization of X-ray projection data with custom calibration phantoms allows accurate measurements to be obtained on the radiodense samples after applying the proposed correction method on biomedical micro-CT images.
