Reliability of Linear and Curvilinear Measurements on Cone- Beam Computed Tomography Images for the Evaluation of Implant Sites and Jaw Pathologies

Objective: To evaluate the intra-examiner and inter-examiner reliability of linear and curvilinear measurements for the complete assessment of implant sites and jaw pathologies using Cone-Beam Computed Tomography (CBCT). Material and Methods: Fifty cone-beam computed tomographic images of patients were retrieved from the archives of Dentomaxillofacial Radiology. CBCT images taken for implant planning and evaluation of intrabony jaw pathologies (benign cyst/tumor) were included. Two expert oral and maxillofacial radiologists analyzed the images independently and made the measurements. The images for implant planning were analyzed for width, the height of the edentulous site, and the qualitative analysis of bone in the region. Jaw pathologies were assessed for linear dimensions and curvilinear measurements. Results: The inter-observer measurement error for implant site analysis ranged from 0.12 to 0.42 mm with almost perfect agreement (ICC: 0.94 to 1). The inter-observer measurement error for jaw pathology was 0.09 to 0.25 mm (ICC: 0.98-1). Curvilinear measurements showed perfect agreement between the observers. The intraobserver reliability for the various parameters used for the assessment of the implant site and jaw pathologies indicated almost perfect agreement. Conclusion: Reliability between the radiologists is high for various measurements on CBCT images taken for implant planning and jaw pathologies.


Introduction
Cone-beam computed tomography (CBCT) imaging provides detailed information by providing a three-dimensional view for the assessment of lesions affecting the maxillofacial region [1]. Measurements made on CBCT images play an important role in clinical situations. These include the evaluation of edentulous sites for implant placement, extent and size of cysts, tumors that affect the jaws, measuring airway spaces, and cephalometric analysis. CBCT software enables the operator to measure the linear and curvilinear dimensions of structures with a high degree of reliability [2,3].
These measurements serve as a guiding tool for the clinicians to make appropriate decisions. It is well established that measurements of edentulous space for implant placement are accurately obtained via CBCT imaging [4][5][6]. Linear measurements on CBCT images are made for assessing the bone height and width for implant placement along with their distances from adjacent vital structures. These measurements are also used for predicting the dimensions of jaw pathologies and envisage their effect on the surrounding structures [7].
Other indications for measurement include pre-implant bone-density assessment, proximity to inferior alveolar nerve canal, and vital structures. However, the accurate reproducibility and reliability of such measurements become crucial in clinical decision-making and appropriate treatment planning.
It ensures consistency in the diagnosis between radiologists and clinicians and helps in the follow-up of patients. Thus, in these situations, the interobserver and intra-observer variations of these measurements become vital. Studies testing the reliability of these measurements have mostly been done on in vitro models [4,8,9]. Literature focussing on the reliability of these measurements using different software is also tested [10,11]. However, studies focusing on the reliability of these measurements on patients' CBCT images are very few [2].
Hence, the purpose of this study was to evaluate the intra-examiner and inter-examiner reliability of linear and curvilinear measurements on CBCT images taken for implant site analysis and assessment of jaw pathologies.

Study Design and Sample
This retrospective, observational, cross-sectional study was conducted on the CBCT images from the archives of Dentomaxillofacial Radiology of Manipal College of Dental Sciences, Manipal Academy of Higher Education, Manipal, India. These CBCT images were acquired from i-CAT 17-19 Imaging System (Imaging Sciences International LLC, USA), and the images were viewed and analyzed through the In Vivo Dental Anatomage software. Images taken to assess implant sites and estimate the extent of jaw pathologies were included in the study irrespective of their field of view. CBCT images with artifacts in the area of interest were excluded from the study.

Data Collection
Sample size estimation was done based on the acceptable reliability (ICC: 0.5) and expected reliability (ICC: 0.8) [12]. The sample size yielded to be a minimum of 22 at the power of 80% and an alpha of 5% for two observations per image. Two investigators (Oral and maxillofacial radiologists) determined the linear measurements of various implant sites and curvilinear measurements of pathologies on the CBCT images at two instances; at baseline (T0) and repeated the measurements two weeks later (T1) to evaluate the intraexaminer error. The image layer on which the measurements would be made was selected by marking the points on the axial section. Subsequent cross-sectional and reconstructed panoramic images were obtained automatically. The cross-section thickness was set as 0.5 mm, and the section interval was set as 1 mm. Both the observers recorded all the measurements in a specially designed proforma.
Measurements for the analysis of edentulous sites for implant placement were performed in three dimensions: the buccolingual width (mm), the height of the edentulous region (mm), and the qualitative assessment of the bone in Hounsfield units (HU). These readings were done on the "Implant screen" of the CBCT software, and measurements were made on the sagittal sections ( Figure 1). These measurements were made using a wired mouse to position the marking tool cursor in the software. The measurements for the width of the alveolar ridge was obtained by marking two points at the level of the alveolar crest from the buccal/labial cortex to the palatal/lingual cortex. The width of the edentulous site was marked at two levels;  Assessment of jaw pathologies was made on the axial and sagittal sections. The image layer on which the measurements were to be made was chosen by scrolling through the axial sections to obtain the specific section that depicts the jaw pathology in its maximum dimension. Curvilinear measurements were obtained by marking multiple points along the outline of the lesion in its maximum dimension ( Figure 2). Then, the buccolingual, mesiodistal, superoinferior, and curvilinear measurements were made. Additionally, the effect of the pathology on the surrounding structures (root resorption, cortical perforation) was recorded as present or absent ( Figure 3).  All the measurements were analyzed for inter and intraobserver reliability.

Results
This study was done on 50 CBCT images of patients whose images were analyzed for possible implant placement and assessment of the extent of jaw pathologies. Implant analysis and measurements were were in the mandibular region.
The mean measurements of the implant site assessment and jaw pathologies, along with the measurement error, are depicted in Table 1. The examiners' overall agreement was almost perfect, with an inter-examiner measurement error of 0.12 mm for the first dimension W1 (ICC=0.94). The agreement was strong even for the second reading W2 with an inter-examiner error of 0.17 (ICC=0.98). There was a perfect agreement with the second dimension (height of the edentulous ridge) with a measurement error of 0.42mm and an ICC value of 1 (Table 2). There was substantial agreement between the observers for the third dimension, i.e., qualitative bone assessment values (Hounsfield units HU) and prediction of favourability of sites for implant placement showed (Kappa=0.62) ( Table 3).  Assessment of jaw pathologies suggested that the various parameters made by the observers suggested almost perfect agreement (ICC values: 0.96 to 1) (Table 4). Additionally, there was a perfect agreement regarding the presence or absence of the effect of the jaw pathologies on surrounding structures (Kappa =1) ( Table 5).
The intraobserver reliability for the various parameters used for the assessment of the implant site and jaw pathologies indicated almost perfect agreement (Tables 2 and 4).

Discussion
CBCT has caused a paradigm shift from conventional planar imaging to 3D imaging in the field of maxillofacial radiology. CBCT plays a vital role in clinical situations like presurgical implant planning, assessment of intraosseous jaw pathologies, airway assessment, and cephalometric analysis in orthodontics [14]. Measurements made on the CBCT images are critical as they dictate treatment procedures, mainly; implant placement [5]. The measurements marked by the Oral and maxillofacial radiologist guides the operating clinician regarding the dimensions of intraosseous pathologies affecting the jaws.
Studies have evaluated the accuracy of CBCT measurements with actual clinical measurements in vitro scenarios and proved that CBCT values are highly reliable and accurate [3,5,8,9,15,16]. The measurements' reliability was excellent in studies on the diagnosis of upper airway disorders [17]. On the other hand, another study testing the reliability of CBCT in upper airway assessment showed that only oropharyngeal volume had excellent intra and interexaminer reliability and hypopharynx and nasopharynx volumes and the overall minimal cross-sectional area had a low agreement between the observers [18]. A systematic review on the reliability and reproducibility of three-dimensional cephalometric landmarks on CBCT showed that landmarks on the median sagittal line and dental landmarks had the highest reliability. In contrast, the landmarks on the condyle, porion, and orbitale presented lower levels of reliability [19].
Golfeshan et al. [20] evaluated the reliability of impacted maxillary canine feature measurements taken by CBCT and concluded that the angulation of the canine had the highest reliability, while the rotation and dilaceration had the lowest reliability.
Previous authors evaluated the reliability of linear measurements on CBCT for the proximity of mandibular third molars to the nerve canal. They found that there was a strong agreement between examiners with significant reliability and reproducibility [21]. A systematic review of the accuracy of measurements on CBCT images related to presurgical implant treatment planning suggested that CBCT can be considered an appropriate diagnostic tool for 3D preoperative planning. Also, it suggested that in clinical practice, the accuracy of measurements on CBCT images is influenced by factors such as patient movement, metallic artifacts, device-specific exposure parameters, and the software used [22].
An in vitro CBCT study utilizing different exposure parameters to test the accuracy of linear measurements around dental implants placed in the maxilla of dry human skulls showed acceptable accuracy for linear measurements around dental implants, despite the exposure parameters used [5]. Prashanti et al.
[23] compared the measurements of implant site planning in 37 patients on ridge mapping, panoramic imaging, CBCT and found a moderate correlation between the measured bone widths from ridge mapping and CBCT. There was a high correlation in the measurements obtained between panoramic imaging and CBCT.
On the other hand, Luk et al. [24] compared ridge mapping and CT for implant planning of twenty-one sites in 14 patients and concluded that the values differed significantly.
In the present study, the overall agreement between the examiners was almost perfect for all three dimensions. The inter-examiner measurement error was 0.12 mm for the first width (W1) and 0.17 mm for W2. Also, the inter-examiner measurement error for the height of the edentulous ridge was 0.42 mm. The values observed from this study showed almost perfect agreement between the observers when compared to the findings from the study by Safi et al. [2].  [25].
CBCT imaging provides information on dental and maxillofacial structures for diagnosis and surgical planning in a multiplane mode (axial, coronal, and sagittal planes). It additionally offers important findings regarding the extent of the lesion, bone resorption, sclerosis of neighbouring bone, cortical expansion, presence of calcifications, and the proximity to other vital anatomical structures [1,14,26].
There are no studies that have tested the reliability of measurements of bone pathologies of the jaws on CBCT images. Also, this is the first-ever study that has tested the reliability of curvilinear measurements on

Conclusion
Hence, this study showed that there was almost perfect inter and intraexaminer reliability in the linear and curvilinear measurements made by the radiologists for implant site analysis and assessment of intrabony jaw pathologies. Minimal measurement differences may be attributed to reasons like the radiologist's skill, changes in the inclination at which measurements are made, and selection of the exact image layer by both the examiners. Future studies should assess the reliability of the measurements on CBCT in various other clinical scenarios. Conceptualization, Software, Writing -Review and Editing and Supervision. All authors declare that they contributed to critical review of intellectual content and approval of the final version to be published.

Financial Support
None.