ESTRO 2024 - Abstract Book

S3742

Physics - Image acquisition and processing

ESTRO 2024

31

Digital Poster

Image quality of Ethos CBCT in radiotherapy in pelvis mode

Erik B van Dieren 1 , Jette J Slettenhaar 2 , Jeroen Veltman 3

1 Medisch Spectrum Twente, Radiotherapy, Enschede, Netherlands. 2 University of Twente, Faculty Science & Technology, Enschede, Netherlands. 3 University of Twente, Multi-Modality Imaging, Enschede, Netherlands

Purpose/Objective:

The quality of Cone Beam CT (CBCT) pelvis images for the Ethos™ (1) is evident, even for non-HyperSight systems, but Catphan CBCTs are very similar to those of a TrueBeam in terms of resolution and low contrast. It is speculated that this is due to the (head) size of this phantom during acquisition. The goal of this study was to investigate CBCT differences between systems using larger, pelvis like phantoms.

Material/Methods:

Pelvis CBCTs were acquired on two Ethos systems, and on two TrueBeams, using a Kyoto Phantom (2) or a 20cm (homogeneous) water phantom. For Ethos systems, both the non-iterative and iterative CBCT were acquired. For the two TrueBeams, only the non-iterative option was available, and analysis results were combined to define a base-line to establish Ethos CBCT quality.

Data were imported in MatLab (3), and analysis was performed using line profiles and regions of interest. Contrast was computed as (HU Prostate-HU background)/(S.D. background). HU accuracy for Kyoto “prostate” and “bone” was computed. In addition, spatial resolution was defined as Modulation Transfer Function @10%, using a the Fourier Transform of a HU profile over “prostate”. Measurements were performed weekly, for 1 year, from which mean and S.D. were calculated. During the study, CBCT QA was performed according to vendor schedule.

Results:

Image quality was quite constant during the study, reflected by the low S.D. (Table 1). However, there is a substantial difference between the 2 Ethos systems : Ethos1 performs better than TrueBeam, and iterative reconstruction improves results, except resolution, which decreases to below that of TrueBeam. Contrarily, for Ethos2, only HU “bones” is better than TrueBeam, the other parameters are worse. Iterative further decreases parameters, except for resolution which is now higher than both Ethos1 and TrueBeam. Analyzing HU profiles, the right side part of the images always had larger HU values than the left side part. This was found for Kyoto in supine, prone, and upside down position, and for scans of the water phantom (fig 1). Better symmetry correlates with better image quality parameters, at least for the studied systems. This asymmetry is absent in ant-post direction for all scans. Also, it is absent in left-right profiles for TrueBeam focal spot acquisition mode, where the imaging panel is centered. Performed CBCT QA showed that systems were within specifications, so no CBCT re-calibration influenced the results of this study.