Rubens Spin-Neto

Standardized method to quantify the variation in voxel value distribution in patient-simulated CBCT datasets

Publikation: Bidrag til tidsskrift/Konferencebidrag i tidsskrift /Bidrag til avisTidsskriftartikelForskningpeer review

Objective: To suggest a standardized method to assess the variation in voxel value distribution in patient-simulated CBCT datasets, and the effect of time between exposures (TBE). Additionally, a measurement of reproducibility, AMORe (Aarhus Measurement of Reproducibility) is introduced, which could be used for quality assurance purposes. Methods: Six CBCT units were tested (Cranex-3D/CRAN; Scanora-3D/SCAN; NewTom-5G/NEW5; i-CAT/ICAT; 3D-Accuitomo FPD80/ACCU; NewTom-VG/NEWV). Two sets of volumetric data of a wax-imbedded dry human skull (containing a titanium implant) were acquired by each CBCT unit at two sessions on separate days. Each session consisted of 21 exposures: one "initial" followed by a 30-minute interval (initial dataset), 10 acquired with 30-minute TBE (datasets 1 to 10), and 10 acquired with 15-minute TBE (datasets 11 to 20). CBCT data were exported as DICOM files and converted to text files containing x-, y-, and z-position and grey shade for each voxel. Subtractions were performed voxel-by-voxel in two set-ups: 1)between two consecutive datasets, and 2)between any subsequent dataset and dataset 1. The mean grey shade variation for each voxel was calculated for each unit/session. Results: The largest mean grey shade variation was found in the subtraction set-up 2 (27-447 shades-of-grey, depending on the unit). Considering subtraction set-up 1, the highest variation was seen for NEW5, between datasets 1 and the initial. Conclusion: Discrepancies in voxel value distribution were found comparing the initial exam of the day with the subsequent exams. TBE had no predictable effect on the variation of CBCT-derived voxel values. AMORe ranged between 0 and 64.

OriginalsprogEngelsk
TidsskriftDentomaxillofacial Radiology
Vol/bind44
Nummer2
Sider (fra-til)20140283
ISSN0250-832X
DOI
StatusUdgivet - 2014

Se relationer på Aarhus Universitet Citationsformater

ID: 82321481