Premium
Dependence of computed tomography volume measurements upon section thickness: An application to human dry skulls
Author(s) -
Sahin Bunyamin,
Mazonakis Michael,
Akan Huseyin,
Kaplan Suleyman,
Bek Yuksel
Publication year - 2008
Publication title -
clinical anatomy
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.667
H-Index - 71
eISSN - 1098-2353
pISSN - 0897-3806
DOI - 10.1002/ca.20664
Subject(s) - coronal plane , section (typography) , medicine , computed tomography , volume (thermodynamics) , tomography , standard deviation , cross section (physics) , nuclear medicine , multislice , mathematics , geometry , anatomy , radiology , statistics , physics , computer science , quantum mechanics , operating system
Estimation of intracranial volume (ICV) using computed tomography (CT) scans has previously been described. However, we were not able to identify a gold standard study that analyzed the effect of section thickness on the estimation of ICV. Therefore, we conducted the present study, scanning five dry skulls in the coronal and axial planes using a multislice CT machine (Toshiba TSX‐101A, Aquilion 16 Slice, Tochigi, Japan). Consecutive sections of variable thicknesses of 2, 3, 5, 7, and 10 mm, respectively, were used to estimate ICV by means of the planimetry method of the Cavalieri principle. All estimations were done by the same observer. However, the estimated volumes did not concur with the actual volumes of the skulls as determined by the fluid displacement technique ( P < 0.05). In fact, results revealed that the section thicknesses created over‐ or under‐projection effects for the estimated volumes. The results were analyzed to reveal the deviation principles of the estimates based on section thickness. Prediction formulas were calculated to estimate the deviation percentage of the ICV depending on section thickness and section plane. Ultimately, the results showed that the effect of section thickness on ICV estimates could not be overlooked, but that the values obtained could be corrected using the proposed prediction formulas presented in this study. Clin. Anat. 21:479–485, 2008. © 2008 Wiley‐Liss, Inc.