Three-dimensional image reconstruction for CCD camera based optical computed tomography scanner

Gel dosimetry is the best tool for three dimensional evaluation of radiation dose. Radiation dose can be measured from the reconstructed images of the gel. The aim of this study was to develop three-dimensional image reconstruction software for a CCD camera based cone beam optical computed tomography (OCT) scanner. In this study FDK reconstruction algorithm was implemented using C language and the quality of the reconstructed images were observed. The projections were acquired using an in-house prototype OCT scanner. The scanner design includes a broad beam light source, a scanning water tank, a turn table, a polyethylene terraphthalate container and a scientific camera for imaging. Multiple projection images were obtained along a circular trajectory for every degree from 1 degree to 180 degree. The objects imaged were a 4 mm diameter opaque rod placed at the centre of the axis of rotation and a 5 mm diameter glass capillary tube with a bore of 1 mm diameter placed 2.4 cm off centre. The reconstruction of 240 slices of each object was completed in 32 minutes. The diameter of central bore of the capillary tube that was measured from reconstructed image was about 0.9 mm giving an error of only 0.1 mm. Refraction and reflection of light at the walls of the container led to some artifacts. The preprocessing of input dataset enabled noise suppression The CNR values of the preprocessed images showed significantly higher values than their non processed counterparts. This study has enabled implementation of three- dimensional image reconstruction software for a CCD camera based cone beam optical computed tomography scanner.