Accuracy of linear measurements using low dose cone beam computed tomography protocol versus direct skull linear measurements: An in vitro study

Introduction

Two-dimensional (2D) imaging techniques have been used in dentistry since 1896. Despite its long clinical success, 2D imaging possesses a number of problems, including superimposition and magnification, which may result in interpretation problems of the images, whether it actually represents the anatomical structures and/or pathological conditions¹.

Lately, dental imaging techniques have advanced with the introduction of tomography. Cone beam computed tomography (CBCT) is the most recently introduced tomography that greatly approximates the accepted standard for three-dimensional (3D) maxillofacial imaging that can guide diagnosis, treatment planning, and follow-up^2,3. CBCT produces accurate images, leading it to be utilized for many dental fields, such as surgical, endodontics, prosthodontics, and orthodontics^4,5.

The radiation dose imparted by a CBCT examination varies as it depends on many variables, such as the type of the CBCT machine, the chosen field of view (FOV), the number of basis images, the mode of exposure (continuous or pulsed), and the exposure parameters used for scanning⁶. Varying the machine’s exposure parameters will result in considerable reductions in radiation dose, which is considered advantageous from a biological point of view. However, theoretically reductions in radiation dose may possibly lead to under sampling artifacts or quantum noise that could adversely affect the diagnostic quality of the images, thus affecting the accuracy of measurements obtained from these images⁶.

This study aimed to compare the accuracy of linear measurements conducted using a low dose CBCT protocol in comparison with direct skull linear measurements.

Methods

The current study was conducted on ten dry human skulls that were obtained from the Anatomy Department, Faculty of Medicine, Cairo University.

The study was approved by the Research Ethics committee of Faculty of Dentistry, Cairo University.

Skull preparation

Using a blue permanent marker, 23 anatomical landmarks were identified on each skull by marking small points representing each landmark, then 12 linear measurements were conducted directly on the skull between these anatomical landmarks using an electronic digital calliper (Allendale Electronics Ltd, Hertfordshire, UK) (Figure 1). These linear measurements are shown in Table 1.

Gutta-percha cones, size 80, were cut into 2 mm rod and were glued over the drawn anatomical landmarks on the skull, to be used as radiopaque radiographic markers. After gutta-percha application, the skulls were covered with block of pink wax of 10-12 mm thickness, which was adapted carefully on the facial surface of the skull from the inferior border of the mandible till above the frontonasal suture for soft tissue simulation^7,8.

Figure 1. Example of direct linear measurements between Or (Rt) and Or (Rt) using digital calliper.

Table 1. Linear measurements.

Linear measurements	Description
ANS-PNS	From the most anterior point on the maxilla at the nasal base to the tip of the posterior nasal spine of the palatine bone
OR F-OR F	From the middle of the inferior border of the right orbital foramen to that of the left side.
Or-Or	From the lower most point on the lower margin of the right orbit to that of the left side.
MORw-MORw	From the middle of the right medial orbital wall to that of the left side.
LORw-LORw	From the middle of right lateral orbital wall to that of the left side.
Or-P	From the lower most point on the lower margin of the right orbit to the highest point on the upper margin of the external auditory meatus of the right side.
ZY F-ZY F	From the middle of the inferior border of the right zygomatic foramen to that of the left side.
GP-GP	From the medial wall of the right greater palatine foramen to that of the left side.
CO-CO	From the most superior posterior point on the head of the mandibular condyle of the right side to that of the left side.
GO-GO	From the middle of the curvature of the angle of the mandible of the right side to that of the left side.
ANT R- ANT R	From the most posterior point on the curve of the right anterior border of the ramus to that of the left side.
Mental F-Mental F	From the inferior border of right mental foramen to that of the left side.

CBCT radiographic examination

CBCT examinations were performed using Planmeca ProMax 3D Mid CBCT unit (Planmeca, Helsinki, Finland).

The skulls were mounted on the machine, and the laser beams were adjusted to centralize the skull within the scanning field. The skulls were then scanned with a low dose protocol of 90 kVp, 7.1 mA, 9 sec, 600 µm voxel size and 20×20 cm FOV.

After scanning each skull, the reconstructed images were viewed on the computer screen using Romexis Viewer 4.4.O.R software. The same linear measurements conducted on the skulls were conducted on CBCT orthogonal images (Figure 2; Table 1).

Figure 2. Cone beam computed tomography sagittal image showing ANS-PNS horizontal linear measurement, measuring 50.24 mm.

The CBCT and linear measurements were conducted by two observers; the first observer repeated the reading two times with a time interval of one month between each reading.

Results

There was no statistically significant difference between the low dose CBCT measurements when compared to direct skull measurements. Using the low dose protocol, mean DE was recorded as 0.83. RDE ranged from 0.8% to 1.9% for almost all measurements (Table 2).

Discussion

For many CBCT machines, it is possible to optimize one or more of the investigated exposure parameters and therefore reduce the patient's radiation dose, while maintaining diagnostic image quality and accuracy for some diagnostic tasks^9,10. Accordingly, the current study aimed to investigate the effect of reducing the mA and exposure time on the accuracy of CBCT linear measurements as compared with a digital calliper.

The results of the present study revealed that the RDE of almost all the measurements ranged from 0.8% to 1.9%, which didn't exceed 5%. This percentage has been considered as clinically acceptable and permissible relative error in the medical field, as reported by Tarazona-Álvarez et al.¹¹ and Rokn et al.¹².

The highest accepted RDE was recorded on the MORw-MORw horizontal linear measurements, 3.4%. The inverse relation between the RDE and the mean of the gold standard of MORw-MORw among all conducted measurements clearly explained its high RDE relative to its low mean of gold standard, which was 21.37%.

The results of Hidalgo et al.¹³ was in accordance with this study as it showed that the coefficient of variation for measurements was between 1.0% and 1.3% using different tube voltage and tube current. They concluded that a low dose protocol of 80 kV and 3 mA could be used for clinical practice, which represented as much as a 50% dose reduction compared with manufacturer’s recommendations, while giving the operator the freedom to adjust the mA by +0.5 mA on the basis of their judgments of the patient’s size.

Further confirmation was obtained from Vasconcelos et al.¹⁴. They concluded that there was no association between the increase in milliamperage and the reliability of the measurements, and recommended the use of low dose protocols when the purpose of the examination is to obtain linear measurements. They added that the 2 mA and 4 mA should be avoided because they could cause degradation to the image and could affect the visualization of the mandibular cortical bone.

In accordance with the current study, Al-Ekrish⁶ results revealed that on decreasing exposure time the reliability and dimensional accuracy of linear measurements for implant site evaluation were not affected.

In conclusion, the results of this study support the idea that decreasing mA and/or exposure time will not affect the accuracy of linear measurements when craniofacial imaging tasks is required.

Data availability