Comprehensive Characterization of Iterative Reconstruction in a Radiation Oncology Setting
R Price1*, I J Chetty2, C Glide-Hurst2, (1) Wayne State University, Detroit, MI (2) Henry Ford Health System, Detroit, MITU-C-103-4 Tuesday 10:30AM - 12:30PM Room: 103
Purpose: Iterative reconstruction (IR) reduces noise, thereby allowing dose reduction in CT while maintaining comparable image quality to filtered back-projection (FBP). We sought to characterize image quality metrics necessary for integration of IR into treatment planning.
Method and Materials: CTs (Brilliance Big Bore v3.6, Philips Healthcare) were acquired of several phantoms using 120 kVp and 25-800 mAs. IR was applied at levels corresponding to noise reduction of 0.89-0.55 with respect to FBP. CT to electron density (CT-ED) curves were generated over all IR levels. Uniformity, spatial and low contrast resolution were quantified via the CATPHAN. Modulation Transfer Functions (MTFs) of a point source module characterized spatial frequency. Contrast to noise ratio (CNR) was quantified. Line profile and region of interest (ROI) analyses were conducted via scripting functions to quantify boundary changes and noise. Contouring differences were quantified via overlap indices (OI) and Dice similarity coefficient (DSC), and dose calculation was assessed. Imaging dose reduction was determined.
Results: No appreciable changes were observed for CT-ED curves between FBP and IR levels (maximum difference ~6.7 HU for bone). For uniformity, differences were <1 HU between FBP and IR. Spatial resolution was well conserved; largest MTF width difference at 50% maximum between FBP and IR levels was 0.1 lp/cm. Profile analysis suggested slight edge deterioration (<3 HU), although OI and DSC ranged from 0.994-0.999. Maximum dosimetric difference was 0.015%. No notable changes in low-contrast detectability were observed. Substantial noise reduction (~45% maximum reduction in signal standard deviation) and CNR gain (maximum >200%) were observed, corresponding to ~70% dose reduction.
Conclusion: IR meets the criteria needed for treatment planning. Slight degradation in object edges was observed, however this did not impact contouring ability or dosimetric parameters. IR substantially reduced imaging dose, and a prospective trial is warranted to optimize settings for each anatomical site.
Funding Support, Disclosures, and Conflict of Interest: The submitting institution holds research agreement with Philips Healthcare.