CBCT Image Quality Assessment Testing Clinically Relevant Volume Orientation and Position

Public Deposited
Analytics

Downloadable Content

Download PDF
Request Version for Screen Reader
Last Modified
  • March 21, 2019
Creator
  • Kurzweg, Brittany
    • Affiliation: School of Dentistry, Oral Pathology Section, Oral and Maxillofacial Radiology Graduate Program
Abstract
  • Introduction and Objectives: Some physical measures of CBCT image quality correlate well with diagnostic image quality. Traditionally, these measures have been assessed in the center in a standard orientation. The purpose of this study was to test whether measures of image quality vary as a function of test tool location, orientation and dose. The second purpose was to determine if there was an association between objective and subjective image quality. Methods: CBCT objective image quality was assessed with one standard and three modified phantoms using five fields of view. The test tool was located at the center of the phantom (standard), at the periphery (Mod1), angled and at the center (Mod2), or angled plane and at the periphery (Mod3). Phantoms were imaged with a Carestream CS 9300 CBCT scanner (Carestream, Rochester, NY), using SDSR (180-250µm voxel/90kVp/64mAs) and LDLR (400µm voxel/85kVp/14.5mAs) for each field-of-view. Contrast-to-noise ratio (CNR) and 10% modulation transfer function (MTF) were assessed in three repeated volumes. Data were analyzed using ANOVA and Tukey’s HSD. Subjective image quality was assessed with a pairwise comparison of anatomical landmarks corresponding to the test tool locations and orientations. Data were analyzed using logistic regression. Results: CNR differed by phantom (p<0.0001) and dose (p<0.0001) for the 8x8 and 17x11 cm FOVs. Mod3 displayed significantly greater CNR than other phantoms. Low dose protocol provided higher CNR. MTF differed only by dose (p<0.0001) for the 8x8, 17x6, and 17x11 cm FOVs. SDSR provided higher MTF. Dose protocol was statistically significant for subjective image quality. Observers preferred images with higher MTF rather than higher CNR. Mod3 was negatively associated with observer preference. The 17x6cm FOV was positively associated with observer preference. Conclusions: CNR improved for a peripherally positioned angled test tool (Mod3). Reduced kVp and larger voxels appear to counteract the effect of reduced mAs producing improved CNR at LDLR. Thus, image quality parameters are different at the center of a CBCT volume when compared to the periphery, depend on the orientation of the object, and vary as a function of kVp and voxel size. Observers preferred images with a higher MTF rather than higher CNR.
Date of publication
Keyword
DOI
Resource type
Rights statement
  • In Copyright
Advisor
  • Ludlow, John B.
  • Ivanovic, Marija
  • Mol, André
Degree
  • Master of Science
Degree granting institution
  • University of North Carolina at Chapel Hill Graduate School
Graduation year
  • 2017
Language

Relations

Parents:

This work has no parents.

Items

Thumbnail Title Date Uploaded Visibility Actions
file details Kurzweg_unc_0153M_16976.pdf 2019-04-09 Public Download
file details PREMIS_Events_Metadata_0_249972b2-787e-4bb3-a824-cc58c02c3d22.txt 2019-04-09 Public Download
file details original_metadata_file_249972b2-787e-4bb3-a824-cc58c02c3d22.xml 2019-04-09 Public Download