Jun Deng PhD, Presenter: Nothing to Disclose

Ravinder Nath PhD, Abstract Co-Author: Nothing to Disclose

To develop an innovative dynamic filter to further reduce the imaging doses in CBCT scan while achieving 4D exposure modulation on varied patient anatomy.

An EGS4/BEAM Monte Carlo code was employed to characterize the photon beams emanated from the dynamic filter in this investigative study. Full phase space simulations were performed on a Varian OBI CBCT imager with the proposed dynamic filter hypothetically installed to replace the current bowtie filter. Field shaping by the dynamic filter was further simulated with a series of component modules JAWS stacking over each other. Phase space particles were first scored on the back surface of the filter with ‘latch’ turned on to track particle interactions, and then analyzed to assess the filter leaf properties. Furthermore, a hypothetical CBCT scan on a patient CT anatomy with the proposed dynamic filter was simulated with a benchmarked Monte Carlo code. The imaging doses received by the various organs-at-risk and the transit images through the patient anatomy were compared between the default bowtie filter and the proposed dynamic filter.

Our preliminary study indicated that a dynamic filter consisting of two banks of 25 aluminum leaves (0.5 cm thick x 2 cm wide x 7 cm long) can be used to create a 5 x 5 leaf matrix capable of modulating the kilo-voltage x-ray exposure to the patients in three dimensions. As the x-ray imager revolves, the patient contour variation can be further compensated with varying leaf thickness and pattern at each projection angle, hence making the dynamic filter essentially a 4D compensator. The photon leakage and scatter from the dynamic filter were found to be less than 4%, causing negligible degradation to the image quality. With the dynamic filter, the imaging doses to the patient could be reduced by 30-50% as compared to the default bowtie filter.

It is feasible to achieve a 4D exposure modulation with the proposed dynamic filter on the current CBCT scanner. Uniform noise property with clinically acceptable image quality can be achieved across all the projection angles, eliminating the under-exposure or over-exposure often encountered in the clinical applications of CBCT scans.

A dynamic filter can effectively compensate for the varied patient anatomy both spatially and temporally, reducing imaging doses to the patients considerably while maintaining uniform noise property.

Deng, J, Nath, R, A Novel Dynamic Filter for Improved CBCT Scan.  Radiological Society of North America 2012 Scientific Assembly and Annual Meeting, November 25 - November 30, 2012 ,Chicago IL. http://archive.rsna.org/2012/12023995.html