Semi-automatic Measurement of Ocular Volume from Facial Computed Tomography and Correlation with Axial Length

Purpose: To measure the ocular volume from facial computed tomography (CT) scans using a semi-automatic computer program, and to analyze possible correlations between the axial length and ocular volume using regression analysis. Methods: Forty eyes from 20 facial CT scans were used to measure the ocular volumes. The cross-sectional ocular areas were calculated using a semi-automatic program based on MATLAB r2009a (MathWorks, Inc., Natick, MA, USA), and the ocular volumes were calculated from serial cross-sectional areas. The axial lengths were measured by A-scan ultrasound. Statistical analysis including regression analysis was used to determine possible correlations between the ocular volumes and axial lengths. Results: The mean ocular volumes measured in males and females were 7.16 +/- 1.80 cm(3) and 7.24 +/- 3.38 cm(3), respectively. The mean axial lengths measured in males and females were 23.47 +/- 0.69 mm and 23.23 +/- 1.64 mm, respectively. There were positive correlations using Pearson's correlation coefficient and the partial correlation coefficient adjusted by axial length. Using regression analysis, the following statistically significant equation was derived: (ocular volume [cm(3)] = 0.0056558 x axial length(3) [mm(3)] -0.1798106 x axial length(2) [mm(2)] + 32.9008570 [p < 0.001, R-2 = 0.384]). Conclusions: The ocular volume measurement tool in this study was noninvasive and very useful, without special equipment. Accurate estimation of ocular volumes by a statistical equation was feasible, and these findings may be helpful in further study of various ocular diseases and in predicting preoperative and postoperative ocular volumes.