ESTRO 2024 - Abstract Book

S1890

Clinical - Mixed sites, palliation

ESTRO 2024

scanned twice, once with the qCBCT protocol and once with the standard clinical CBCT protocol. Both scans were acquired using the same CBCT acquisition protocols and imaging dose. Subsequently, qCBCT image data was processed using an in-house developed data correction and reconstruction chain. Standard clinical CBCT image quality was evaluated without further data correction processing. An observer-based image quality evaluation study was performed using the qCBCT and clinical CBCT images of the first 10 study participants enrolled into the trial. Eight expert observers, which consisted of 5 radiation oncologists, 2 radiation therapists, and one physicist, were asked to compare the data set images to answer the question “Which CBCT image provides higher image quality?”. Each qCBCT and clinical CBCT image pair was fused such that the same anatomical region was shown in both images. Image pairs were displayed using the same window/level and magnification settings. Each observer ranked the qCBCT and clinical CBCT images in a blinded setting. All 10 participants’ images were for the pelvis and lower abdomen regions. A total of 80 image quality evaluations were performed by 8 observers for 10 participants’ qCBCT -clinical CBCT scan pairs.

Results:

CBCT image quality ranked higher than the standard clinical CBCT in 85% of all observer evaluations. (p value<0.01). The effect of motion artifacts on observers’ rankings was also evaluated (Fig. 2). For CBCT images where motion artifacts were deemed to be low to moderate (8/10 data sets), qCBCT image quality ranked higher than standard clinical CBCT in 94% of the evaluations. For the two data sets where motion artifacts were considered severe, quality rankings of qCBCT and standard clinical CBCT were equal. In visual evaluations, qCBCT was of greater value in patients with higher BMI.