Study On The Magnification Of Liner Measurement On CBCT Images

In recent years, the rapid development of digital technology led to great upgrading in dental technology and comprehensive changes on oral health care equipment. Dental prosthesis processing and implants implanting increasingly tend to automation and precision, meanwhile the power that prosthesis putting on the support organization had caught more and more attention. Dental prosthesis intelligent processing and precision fabrication of surgical template for guided implant placement, such as CAD/CAM digital technology, "digital impression", "3D printing" technology, need refer to CT or CBCT datas; in modeling, three-dimensional finite element model of dental prosthesis, soft and hard dental tissues based on CT anatomical data would make the analysis of oral biomechanics more intuitive; in surveying, the digital technology was often used in craniofacial precision measurements including bone density, bone defects and the transform from a traditional two-dimensional measurement into a three-dimensional measurement. The image data acquires by CT, CBCT technology was the foundation of dental prosthesis intelligent processing and implants precise implanting, if the data was not accurate, the reliability of conclusions would be greatly influenced. Therefor, it was extremely important to make clear the difference between CBCT data and real data.This study was divided into two Parts, in part I: first,we used experimental mould, which had 72 radiopaque lines on its observation surface, to make clear the differencebetween CBCT measurement and physical measurement, and then analyzed regularities of liner measurement magnification at three direction. At last, establish regression equation to estimate magnification of liner measurement on CBCT images. In part II: by measuring the length of extracted teeth, we compared differences between the length of teeth in CBCT images and actual length as well as the length of correction and actual length, to find out whether there was a vertical amplifier error on CBCT images and explore the effect of regression equation on Z-axis to reduce linear measurement error on CBCT images. Thus, we validated the accuracy of the regression equation, which was used with R-value to speculate the magnification of linear measurement on CBCT images, to provide more accurate data sources for dental prosthesis intelligent processing, precise implanting, and three-dimensional finite element modeling.Part I: the experimental mould was a square polymethylmethacrylate plate with a side length of 120 mm and a thickness of 20 mm, whose observation surface contained 72 grooves which have 10 mm length and 0.3mm depth filling with developer to form radiopaque lines, prepared by the CNC machine. The lines parallel with the Coronal plane were defined as X-axis lines, parallel with Sagittal plane were defined as Y-axis lines and perpendicular to the horizontal plane were defined as Z-axis lines. Every two axis divided the plane into four quadrants and each quadrant contained two axial lines, 9 lines on each axis(cross-over design, the distance from dead center to Sagittal, Coronal and horizontal plane were 10 mm, 30 mm, 50mm).Make clear the center of FOV on CBCT images and then put the mould into FOV to make sure the horizontal central plane(HP) and coronal central plane(CP). Put the mould upwards or downwards HP, the reference plane, for 20 mm and 40 mm and then scanned to get images of HP, OP20, OP40, UP20, UP40 with X-axis and Y-axis lines. Put the mould anterior or posterior to the coronal central plane for 20 mm and 40 mm and then scanned to get images of CP, BP20, BP40, AP20, AP40 with Z-axis lines. The distance that radiopaque lines departing from theFOV center was defined as R. The length of each radiopaque line was measured 6 times, and then the magnification was calculated to make the regularity clear.Part II: use Galileo CBCT to measure the length of 32 extracted teeth. First, fix 32 teeth on Arch-formed base plate vertically along the long axis of the tooth in the shape of arch form and in accordance with the order of incisor teeth- canines- premolars. And then make sure the position of teeth distributed symmetrically at anterior, posterior, left, right direction. Finally, fix the base plate wax on the observation plane of polymethylmethacrylate plate. In MPR function interface of Galaxis3 D software, adjust sectional ?images to find the measurement plane, which contained the cusp, root apex and FOV center. Measuring the length of experimental tooth as CBCT measurement(tooth length was the distance from the cusp to root) Measuring the distance from FOV center to the midpoint of the cusp and root as R-value. Put R-value into the regression equation to calculate magnification, thereby correcting the CBCT measurement to obtain correction value. At last, use a slide caliper to measure the distance from tip to the root of experimental teeth as actual value. Normality test was used to examine CBCT measurements, correction value and the actual value whether consistent with normal distribution; paired t test was used to compare whether there were significant difference between the CBCT measurements and actual value, correction value and actual value as well as M1 and M2.The results were as follows: 1. Galileo CBCT measurements of radiopaque lines were significantly greater than physical measurements. The maximum magnification of X-axis radiopaque lines was 3.4% with R-value of 78.3mm and the minimum magnification was-0.4% with R-value of 11.2mm. The maximum magnification of Y-axis radiopaque lines was 3.3% with R-value of 78.3mm and the minimum magnification was-0.6% with R-value of 11.2mm. The maximum magnification of Z-axis radiopaque lines was3.3% with R-value of 78.3mm and the minimum magnification was-0.4% with R-value of 11.2mm. The means of magnification for X-, Y-, Z-axis radiopaque lines were 1.2%, 1.1%, 1.1%. 2. There was no significant difference in the magnification of liner measurement on CBCT images for radiopaque lines symmetrically distributed(P>0.05). 3. Among different planes, the comparison of the magnification for the same line was that: the magnification of the liner images staying away from the center plane was significantly larger than those close to the center plane(P < 0.05) and the magnification of different lines in the comparison among the stop lines was that: the magnification of the lines staying away from the center of the plane was significantly larger than those close to the center(P<0.05). 4. The comparison of the magnification at different direction with the same R-value was that: X-axis was greater than Y- and Z-axis(P<0.05), while Y-axis and Z-axis magnification was closer(P>0.05). 5. Magnification was positively associated with R(rx=0.996,Px<0.001,ry=0.998,Py<0.001;rz=0.997,Pz<0.001), and the regression equation was as follows: Mx=0.0007R2-0.0177R+0.0343 My=0.0008R2-0.0225R+0.0499 Mz=0.0008R2-0.0238R+0.0689. 6. CBCT measurements of experimental teeth were larger than actual value and the difference was significant(P<0.05). The mean and standard deviation of the difference was 0.15±0.067 mm. 95% confidence interval of difference was from 0.13 mm to 0.21 mm. The average percentage error was 0.74%±0.33% 7. There was no significant difference between the correction value and actual value(P>0.05). The mean and standard deviation of the difference was 0.02±0.074 mm. 95% confidence interval of difference was from-0.01 mm to 0.05 mm. The average percentage error was 0.10%±0.36%.8. There was no significant difference between the magnification estimated by R-value and the actual magnification(P>0.05).Based on the above results, conclusions were as follows: 1. Magnification on CBCT images had features of symmetry and regionalism 2. X-axis magnification was greater than Y- axis or Z-axis magnification(P <0.05) 3. There was a linear amplification on cone-beam CT images, and its magnification was positively associated with the distance from the entity to the FOV center.When R equalled to 11.2mm, the magnification was almost 0, however, When R equaled to 78.3mm, the magnification was up to 3.4% 4. There was a vertical amplifier error on CBCT images of teeth, the average percentage vertical amplifier error was 7.4‰, however, by R-value and the correction equation on Z-axis We can reduce the average percentage vertical amplifier error from 7.4‰ to 1.0‰ to acquire closer real size of teeth.