Assessment of 2D and 3D image characteristics of an optical CT scanner for 3D dosimetry

Gel dosimetry is a method of measuring a dose distribution in three dimensions and requires an efficient accurate read-out approach. Optical CT has the potential to provide this. It is the aim of the present study to characterise the components and properties of a commercially available cone beam optical CT scanner and investigate procedures which could be used for commissioning or as part of an ongoing scanner quality assurance program. Light spectra, CCD response, warm-up time, spatial resolution, response to light intensity, uniformity and the influence of scatter were all demonstrated to influence the accuracy of the optical CT scanner, with the light source/ filter combination and warm up time demonstrating the most significant impact. Individual phantom attenuation coefficients were characterised and used to determine the stability of the phantoms over a period of time. Profiles measured from 2D projections and 3D reconstructions were measured and compared to absorbing solutions to determine uniformity and scattering properties. Recommendations are made for a set of quality assurance procedures, which include geometry calibration of the scanner, characterising the light source, verifying the CCD is functioning correctly, checking the reconstructed accuracy and verifying the reconstruction spatial fidelity. Commissioning of an optical CT scanner should include an optimisation of the scanner geometry and lens settings, measurement of the spectra and temporal stability of the light source, measurement of CCD reciprocity, quantification of attenuation coefficient accuracy, uniformity and noise and measurement of spatial resolution. This work has shown that appropriate warm up times, light source and filter combinations and phantom material selection and manufacturing processors can significantly influence measured attenuation coefficient accuracy, uniformity and noise.