Trial Outcomes & Findings for Evaluation of a Cone-beam CT Scanner for Image Guided Radiotherapy (NCT NCT05524454)

Results Overview

For this objective, the image noise level was evaluated in regions of interest representing different tissue types (muscle and fat) in both CT simulation images and HyperSight CBCT images. Image noise is the variation in image intensity, which is measured in Hounsfield Units (HU), in a region that should have uniform intensity without the presence of noise. Image noise is measured as the standard deviation of HU within the defined regions of interest. Higher noise values indicate poorer image quality. HyperSight images can be reconstructed from raw data using different reconstruction algorithms. For this analysis, both the standard Feldkamp-Davis-Kress (FDK) algorithm and an iterative metal artifact reduction (MAR) algorithm were evaluated.

Recruitment status

COMPLETED

Study phase

NA

Target enrollment

41 participants

Primary outcome timeframe

Within 1 month from the start of treatment

Results posted on

2026-01-06

Participant Flow

Participant milestones

Participant milestones
Measure	Novel CBCT Imaging All subjects undergo one additional imaging session with the novel CBCT imaging system.
Overall Study STARTED	41
Overall Study COMPLETED	40
Overall Study NOT COMPLETED	1

Reasons for withdrawal

Reasons for withdrawal
Measure	Novel CBCT Imaging All subjects undergo one additional imaging session with the novel CBCT imaging system.
Overall Study Withdrawal by Subject	1

Baseline Characteristics

Race and Ethnicity were not collected from any participant.

Baseline characteristics by cohort

Baseline characteristics by cohort
Measure	Novel CBCT Imaging n=40 Participants All subjects undergo one additional imaging session with the novel CBCT imaging system.
Age, Continuous	67 years n=40 Participants
Sex: Female, Male Female	22 Participants n=40 Participants
Sex: Female, Male Male	18 Participants n=40 Participants
Body Mass Index (BMI)	26.4 kg/m^2 n=38 Participants • Height and/or weight values were not available for 2 study participants.

PRIMARY outcome

Timeframe: Within 1 month from the start of treatment

Population: Images were analyzed according to the anatomical region being imaged. As of the last date of publication to ClinicalTrials.gov, only data for the 7 Head/Neck participants were available.

For this objective, the image noise level was evaluated in regions of interest representing different tissue types (muscle and fat) in both CT simulation images and HyperSight CBCT images. Image noise is the variation in image intensity, which is measured in Hounsfield Units (HU), in a region that should have uniform intensity without the presence of noise. Image noise is measured as the standard deviation of HU within the defined regions of interest. Higher noise values indicate poorer image quality. HyperSight images can be reconstructed from raw data using different reconstruction algorithms. For this analysis, both the standard Feldkamp-Davis-Kress (FDK) algorithm and an iterative metal artifact reduction (MAR) algorithm were evaluated.

Outcome measures

Outcome measures
Measure	Head/Neck Imaging n=7 Participants Evaluation of images of study participants treated for head/neck cancer.
Image Quality Noise in fat for HyperSight FDK	7.6 standard deviation of HU values Interval 3.1 to 23.0
Image Quality Noise in muscle for planning CT	6.6 standard deviation of HU values Interval 4.2 to 8.3
Image Quality Noise in muscle for HyperSight MAR	9.8 standard deviation of HU values Interval 3.9 to 14.9
Image Quality Noise in muscle for HyperSight FDK	11.2 standard deviation of HU values Interval 8.8 to 27.4
Image Quality Noise in fat for planning CT	8.9 standard deviation of HU values Interval 4.6 to 13.1
Image Quality Noise in fat for HyperSight MAR	7.1 standard deviation of HU values Interval 3.6 to 15.3

SECONDARY outcome

Timeframe: Within 1 month from the start of treatment

Population: Dose calculation accuracy was analyzed according to the anatomical region being imaged. As of the last date of publication to ClinicalTrials.gov, only data for the 7 Head/Neck participants were available.

The accuracy of dose calculation on HyperSight CBCT was evaluated by re-calculating a participant's clinical treatment plan on one of their HyperSight images and comparing the doses delivered to target structures and organs at risk (OARs). The lower the difference in dose metrics between a HyperSight CBCT-based plan and a CT simulation-based plan, the more accurate the dose calculation on HyperSight CBCT. Both the maximum dose delivered to OARs and the mean dose delivered to OARs were compared between HyperSight-based plans and CT simulation-based plans. We report the worst overall accuracy, i.e. maximum difference in dose metrics across all OARs from across all patients. Similarly, the maximum dose delivered to the target, the mean dose delivered to the target and the minimum dose delivered to 95% of the target were compared between HyperSight CBCT-based plans and CT simulation-based plans. We report here the maximum difference across all dose metrics from across all patients.

Outcome measures

Outcome measures
Measure	Head/Neck Imaging n=7 Participants Evaluation of images of study participants treated for head/neck cancer.
Accuracy of Radiation Dose Calculation on HyperSight CBCT Max difference in OAR DVH metrics between HyperSight MAR and CT simulation across all subjects.	1.1 Gy
Accuracy of Radiation Dose Calculation on HyperSight CBCT Max difference in target DVH metrics between HyperSight MAR and CT simulation across all subjects.	6.5 Gy

SECONDARY outcome

Timeframe: Within 1 month from the start of treatment

Population: Images were assessed according to the anatomical region being imaged. As of the last date of publication to ClinicalTrials.gov, only data for the 7 Head/Neck participants were available. Seven observers evaluated two image series - CT simulation images and HyperSight CBCT - from each study participant.

CT simulation images and HyperSight CBCTs from subjects were randomly presented to expert reviewers, who were blinded to the type of image they were looking at and who were asked to rate their confidence level for contouring organs at risk (OARs) on the images using a 5-point Likert scale (very high, high, moderate, low, very low). The percentage of times that a reviewer rated their confidence level for contouring OARs on HyperSight CBCT as moderate, high, or very high was recorded. A higher percentage means a more favorable assessment for OAR contouring on HyperSight CBCT. The percentage of times the reviewer's confidence level for contouring on a HyperSight CBCT was equal to or higher than their confidence level on the corresponding CT simulation image was also recorded, where again a higher percentage means a better assessment for OAR contouring on HyperSight CBCT. The HyperSight iterative metal artifact reconstruction (MAR) algorithm was used in this analysis.

Outcome measures

Outcome measures
Measure	Head/Neck Imaging n=98 Evaluations Evaluation of images of study participants treated for head/neck cancer.
Qualitative Assessment of Novel CBCT Images for Contouring Organs at Risk % evaluations OAR contouring confidence on HyperSight MAR was equal or higher than on CT simulation	47 % evaluations
Qualitative Assessment of Novel CBCT Images for Contouring Organs at Risk Percentage of moderate, high, or very high confidence in contouring OARs on HyperSight	70 % evaluations

Adverse Events

Novel CBCT Imaging

Serious events: 0 serious events

Other events: 0 other events

Deaths: 0 deaths

Serious adverse events

Adverse event data not reported

Other adverse events

Adverse event data not reported

Additional Information

Sean Davidson

Varian Medical Systems

Phone: 437-991-8294

Email: [email protected]

Results disclosure agreements

Principal investigator is a sponsor employee
Publication restrictions are in place