Development of a Dual-Energy Cone Beam CT Technology on Ethos Hypersight
Abstract
Purpose
Dual-energy CT (DECT) enables material differentiation and artifact reduction that can improve image quality for radiation therapy treatment planning. Combining DECT with Ethos HyperSight CBCT, which incorporates automatic scatter correction, has the potential to extend these benefits to on-board imaging with HU accuracy approaching conventional CT.
Methods
We performed several phantom studies to test these capabilities. Two calibration studies were performed. First, inserts of iodine (up to 62 mg/mL) and calcium (up to 375 mg/mL) were scanned at low (80 kVp) and high (140 kVp) energies. Second, ten bone- and tissue-equivalent inserts were scanned with the same parameters. A dual-energy (DE) material specific imaging algorithm was developed utilizing DE Hounsfield Unit (HU) values to separate material groups via calculation of the Mahalanobis distance to each groups mean HU to determine the lowest concentrations or densities at which the materials can be reliably differentiated with 95% confidence. The algorithm’s performance was then evaluated on an anthropomorphic phantom. Additionally, a set of Virtual Monoenergetic Images (VMIs) were developed using a separate algorithm to investigate the contrast-to-noise ratio (CNR) improvement at keV energies compared to a clinical head (100 kVp, 88 mAs) imaging protocol.
Results
The minimum concentration at which iodine and calcium can be separated with 95% confidence is 45 mg/mL and 300 mg/mL, respectively. Material specific images were successfully developed with the ability to separate six groups of materials (lung, fat tissue, soft tissue, bone, iodine, and calcium) with 95% confidence. Iodine and bone can be reliably differentiated above 25 mg/mL and 1.4 g/cm3, respectively. VMIs improve CNR by 52% for inner bone, 60% for general adipose and 88% for exhale lung.
Conclusion
We have developed dual-energy CBCT technology on the Ethos HyperSight platform which has the potential to improve image quality and streamline treatment planning workflows in radiation therapy.