Related Experiment Video
Updated: Oct 22, 2026
Reliability of Artificial Intelligence-Based Cone Beam Computed Tomography Integration with Digital Dental Images
Published on: February 23, 2024
Image quality robustness of HyperSight-enabled versus conventional iterative cone-beam CT across patient size and
Nicholas Nelson1, Sorour Hosseini1, Geoff Nelson1
1Department of Radiation Oncology, University of Utah, Salt Lake City, Utah, USA.
Background:
Recent advances in cone-beam computed tomography (CBCT) have led to the development of next-generation CBCT systems with enhanced image quality and quantitative imaging capabilities. These advances are particularly relevant in the context adaptive radiation therapy, where improved soft tissue visualization and Hounsfield unit (HU) stability are increasingly important for target localization and CBCT-based dose calculation.
Purpose:
To determine how image quality and HU stability change with increasing phantom size, non-cylindrical geometries, and exposure level, and compare the robustness of Acuros-based HyperSight-enabled iterative CBCT (HS-iCBCT) and conventional iterative CBCT on a TrueBeam (TB-iCBCT) under these conditions.
Methods:
The CatPhan® 604 phantom, combined with custom annular extensions, was used to simulate increasing patient size and non-cylindrical imaging geometries. Three configurations (no annulus, small annulus, and large annulus) were imaged across a range of exposure levels (mAs) using HS-iCBCT and TB-iCBCT. Image evaluation metrics included noise, contrast-to-noise ratio (CNR), low-contrast visibility (LCV), image uniformity, spatial resolution, and Hounsfield unit (HU) constancy.
Results:
For the no-annulus configuration, image quality was comparable between systems, with slightly lower noise observed for TB-iCBCT. With increasing phantom size through the added annuli, TB-iCBCT exhibited substantial degradation in noise, contrast, and uniformity, whereas HS-iCBCT maintained more consistent performance. Spatial resolution was primarily dependent on field of view and was similar between systems, with smaller fields of view provided improved spatial resolution. HU constancy degraded with increasing phantom size and material density for both systems but remained substantially more stable for HS-iCBCT.
Conclusions:
HS-iCBCT demonstrated improved robustness to patient size and imaging geometry, providing superior image quality and HU stability compared to TB-iCBCT. These improvements may support potential lower-dose imaging and more reliable performance in larger patients and non-cylindrical imaging geometries commonly encountered in clinical radiotherapy.
