(Thursday, 7/18/2019) 10:00 AM - 12:00 PM
Room: 304ABC
Purpose: Rather than representing patient dose, CTDI provides a standardized approach to quantifying radiation output of a given CT scanner and protocol. In 1995, the concept of CTDIweighted was proposed to reflect the average dose absorbed in a section in a cylinder. The associated weighting factors were derived based on the assumption that CTDI(r) is linearly related to the radial location r. The purpose of this work was to re-examine the validity of this assumption for modern CT systems.
Methods: A pencil chamber was suspended in an in-house 16 cm cylindrical water tank. The position of the chamber was controlled by two orthogonal translation stages, so that the chamber can reach any arbitrary axial location. The phantom was scanned by an MDCT using either a large or small bowtie. Scans were also performed with the phantom 6cm off-centered. Under each condition, CTDI100 was measured at 29 axial locations.
Results: Even with the phantom at iso-center, the small bowtie led to a radial profile that is not linear and rolls off at large r; the profile was found to follow a r*cosine(r) function with R^2 of 0.99. As soon as the phantom is off-centered, the dose profile is no longer linear and can be modeled using a second-order 2D polynomial function. The corresponding R^2 was 0.99. Compared with the CTDIweighted provided by fitting data measured at iso to the linear model (22.2 mGy), the actual value was 17.5 mGy at off-centered location. Collimation width did not exert any significant impact on the shape of the axial dose profile.
Conclusion: Due to the use of bowtie filters in modern CT, the weighting scheme used in CTDIweighted may need to be revised, and radiation output at locations away from the isocenter may no longer be accurately represented by CTDIweighted measured at iso.
Keywords
Not Applicable / None Entered.
Taxonomy
Not Applicable / None Entered.
Contact Email
