Evaluation of reliability and reproducibility of linear measurements of cone-beam-computed tomography :[PAUTHORS], Indian Journal of Dental Research (IJDR)

ORIGINAL RESEARCH
Year : 2012 | Volume : 23 | Issue : 4 | Page : 473--478

Evaluation of reliability and reproducibility of linear measurements of cone-beam-computed tomography

Thiago de Santana Santos¹, Ana Clďż˝udia Amorim Gomes², Daniela Guimarďż˝es de Melo², Auremir Rocha Melo², Josuel Raimundo Cavalcante², Lorena Cďż˝ssia Gueiros de Araďż˝jo², Rosana Maria Coelho Travassos², Paulo Ricardo Saquete Martins-Filho³, Marta Rabelo Piva³, Heitor Fontes Silva³,
¹ Ribeirďż˝o Preto Dentistry College (FORP/USP), Oral and Maxillofacial Surgery Post-Graduation Student. Ribeirďż˝o Preto, SP, Brazil
² Pernambuco Dentistry College (FOP/UPE), Camaragibe, PE, Brazil
³ Sergipe Federal University (UFS), Aracaju, SE, Brazil

Correspondence Address:
Thiago de Santana Santos
Ribeirďż˝o Preto Dentistry College (FORP/USP), Oral and Maxillofacial Surgery Post-Graduation Student. Ribeirďż˝o Preto, SP
Brazil

Abstract

Objective: it was to evaluate intra- and interexaminer reliability and reproducibility of linear measurements in cone-beam tomography (CBCT) images. Materials and Methods: CBCT scans were obtained from 50 patients of a private clinic. Two examiners located the landmarks twice in two-dimensional and three-dimensional images on different days, with images performed 1 week apart. Intra- and interexaminer reproducibility and reliability were analyzed. Results: the measurement error (ME) did not show significant differences between intra- and interexaminers. The intraclass correlations (ICC) between the intra- and interexaminer readings obtained with the different methods showed almost perfect matches. The results demonstrated high examiner reproducibility for linear and volumetric parameters with high intraclass correlation coefficient (ICC) and coefficient of variation (CV). The ICC showed that the methodology was highly reliable and reproducible (ICC- >0.99 and CV <1.5%). Conclusion: the measurements demonstrated strong agreement between examiners and significant reliability and reproducibility. Therefore, this methodology can serve as a standard for linear measurement analysis of the topography of mandibular canal and osseous adjacent structures.

How to cite this article:
Santos T, Gomes AA, de Melo DG, Melo AR, Cavalcante JR, de Araďż˝jo LG, Travassos RC, Martins-Filho PR, Piva MR, Silva HF. Evaluation of reliability and reproducibility of linear measurements of cone-beam-computed tomography.Indian J Dent Res 2012;23:473-478

How to cite this URL:
Santos T, Gomes AA, de Melo DG, Melo AR, Cavalcante JR, de Araďż˝jo LG, Travassos RC, Martins-Filho PR, Piva MR, Silva HF. Evaluation of reliability and reproducibility of linear measurements of cone-beam-computed tomography. Indian J Dent Res [serial online] 2012 [cited 2023 Sep 28 ];23:473-478
Available from: https://www.ijdr.in/text.asp?2012/23/4/473/104952

Full Text

The knowledge of morphology and topography of mandibular canal for oral and maxillofacial surgery in mandible is important to grasp due to necessity of mandibular nerve preservation. [1] If damage occurs, sensorial alteration as paresthesia and anesthesia of the lower lip and perioral region would be present. [2] The most frequent cause is partial or complete severance of the nerve during the use of rotating or other instruments without seeing. [3] For damaging prevent, it is necessary to estimate the proximity of the mandibular canal to adjacent structures with imaging exams, mainly for removing the third molars [4] as well as for implants placement. [5]

Panoramic radiographs are widely used in dental practice for the diagnosis of many dental pathologies; [6] however, it has the lack of three-dimensional (3D) information of the imaged area needed for optimal preoperative planning of dental implant placement. [7] On the other hand the computed tomography (CT) provides clinicians with the ability to investigate the inner depths of the human anatomy slice by slice through computer reformation of radiographic images. Since the invention of the first CT scanner by Godfrey Hounsfield of Britain in 1972, CT scanners have gained sophistication and have been used in an increasing range of clinical applications. With the recent introduction of cone-beam CT (CBCT) specifically designed for the volumetric imaging of the maxillofacial area, the radiation dose to the patient for a maxillofacial study has been significantly reduced. [8],[9] Because conventional medical CT technology differs from cone-beam computed tomography (CBCT) in the choice of x-ray sources, detectors, and reconstruction algorithms, performance characteristics may differ. [10]

High repeatability and reproducibility of such measurements are important in clinical practice. This is essential in evaluating consistency of diagnosis between different observers and follow-up of patients over a period of time. The correlations between intraexaminer and interexaminer proved to be a positive method in several studies of the CBCT, as well as available software made precise measurements performed by different professionals. [5],[11] However, in daily practice, it is noted that often what the radiologist mark in the printed films is not the accurately measured region for implant surgery. Thus, making necessary the use of digital imaging or millimeter rules for planning measures before installing dental implants. Therefore, the aim of this study was to evaluate intraexaminer and interexaminer reliability and reproducibility of linear measurements in cone beam tomography images for methodology validation.

Materials and Methods

CBCT scans were obtained from 50 patients of a private clinic that third molar removal was intended. Both sides of the images were selected; thus, one-hundred sides were evaluated for linear measurements.

For sample selection criterion was established that all the patients must be adults with third molars present bilaterally and tomographic signs of close relationship of these teeth with the image of the mandibular canal. Exclusion criteria were patients with third molars with symptoms and/or CT signs outside the normal pattern.

CBCT scans were taken using the i-CAT Classic (Imaging Sciences International, Philadelphia, USA) at 120 kV, 7.0 mAs, and 6 mm aluminum filtration. Each image was obtained from 360 slices and was converted to the DICOM format with i-CAT software. When i-CAT software (Imaging Sciences International, Hatfield, USA) was used, the DICOM format images were rendered into a volumetric image. Sagittal, axial, and coronal volumetric slices, as well 3D reconstruction of the image, were used to determine landmarks positions. Computed tomography data were imported into i-CAT software for primary reconstruction of images. Primary reconstructions were made using the large window setting with 0.25 mm axial slices. Axial slicing was oriented parallel to the Frankfort plane.

Secondary reconstruction and measurement of the CBCT images was accomplished using the i-CAT software package. With operator-selected 0.25-mm slice thicknesses, the panoramic tool was used on a representative axial slice to reconstruct the CBCT data in the form of a panoramic image. Initial mapping of the panoramic path was adjusted as needed to ensure that all anatomical points of interest could be easily located within the slice. Once the reconstruction had taken place, axial views of the data were used to locate specific anatomical four points in each side. The measurements 1 and 2 were encountered by writing a straight tangent in ascendant ramus of mandible to down, and two distances were measured from external oblique line and basilar of the mandible until superior and inferior aspect of mandible canal, respectively [Figure 1].{Figure 1}

Measurements were made using a computer mouse to position the two-dimensional (2D) measurement tool cursor at one and then the other end of the landmarks of interest. The distance between landmarks along the curved plane at the center of the panoramic image layer was automatically calculated and displayed. Lengths of the vertical markers distances were measured for each side.

The i-CAT CBCT software allows the user to measure linear distances in 3D by using two different axial slices. Vertical and horizontal markers were measured by mouse clicking the three-dimensional measurement tool on the marker in the first axial slice in which it appeared, as well as the last axial slice before it disappeared. The operator then scrolled through the stack of axial slices until the mark for the other landmark was located. The measurements 1 and 2 were encountered by writing two straight tangents in alveolar bone crest and basilar of mandible, respectively. From these, the distance to mental foramen was measured [Figure 2]. After selecting these landmarks, the linear distance between the points was automatically calculated. [Figure 2] provides examples of sagittal slice containing measurements 3 and 4 in the mental foramen region. Measured distance and pairs of crosshairs indicating the trajectory between points are automatically recorded on the images.{Figure 2}

Both investigators located the landmarks twice in two-dimensional and three-dimensional images on different days with the same hour for equivalent luminosity and the same computer display and contrast configuration, with images performed 1 week apart. It was suggested that they stop once they were feeling tired and continue another day, to reduce exhaustion effect. A description and a definition for each landmark used are given in [Figure 1] and [Figure 2]. Examiner A was an experienced oral and maxillofacial radiologist. Observer B was an experienced oral and maxillofacial surgeon.

Intraexaminer and interexaminer reliability and reproducibility values were determined using the intraclass correlation coefficient (ICC). Values for ICC range from 0 to 1. When ICCs are closer to 1, reliability and reproducibility are stronger; ICC values >0.75 show good reliability. [12] The values were calculated with single measurements from a two-way random model and absolute agreement type of ICC in a statistical software package (SPSS v.13, SPSS, Chicago, IL). To assist in the interpretation of the clinical significance of landmark identification differences, average mean differences (landmark identification error) between examiners (two examiners) were summarized, and descriptive statistics was applied. Coefficient of variation (CV % = [SD/its mean value] × 100) can describe the reproducibility of measurement as a precision error. [12]

Interexaminer variation was determined by paired t-tests for skull, two-dimensional, and three-dimensional measurements. The Kolmogorov-Smirnov test was used to determine normal or heterogeneous data distribution and appropriate parametric or nonparametric tests were then employed. Intraexaminer measurement variation was determined by paired t-tests of first and second measurements. To compensate for multiple comparisons, significance was set at α = 0.01.

Results

The age of patients ranged from 17 to 36 years, with an average of 24.92 years, standard deviation of 4.72 years and a median of 24.50 years. More than half of the patients (60.0%) were aged 17-25 years and 40.0% had 26 to 36 years, the gender distribution was approximately equal being 52.0% of females.

The overall agreement between the two examiners was satisfactory. Differences in percentage were observed between the intra- and interexaminer linear measurements performed varying 1.15 to 7.48. The differences were higher in measure 3 for evaluation and examiners [Table 1].{Table 1}

The measurement error (ME) did not show significant differences between intra- and interexaminers. The ICC showed that the methodology was highly reliable and reproducible (ICC >0.99 and CV <1.5%). The ICCs were generally high (ICC range: 0.99-1.00) by side evaluated according to examiner and time of evaluation. The intraclass correlations (ICC) between the intraexaminer and interexaminer readings obtained with the different methods showed almost perfect matches. The results demonstrated high examiner reproducibility for linear and volumetric parameters with high intraclass correlation coefficient (ICC) and coefficient of variation (CV). Average measurement errors ranged from 0.01 mm (measurement 1 right) to 0.7 mm (measurement 3 right) [Table 2]. {Table 2}

In both intraexaminer and interexaminer, the Kolmogorov-Smirnov test did not reject the similarity of the frequency distribution; thus, the paired t-test was used to compare the mean and standard deviation of proportions by side according to the examiner and the time of evaluation. Significant difference (P > 0.05) was observed in the right side for intraexaminer 1. Moreover, the mean measurements and standard deviation of proportions by side according to the examiner and the time of evaluation were significantly different in measurements 3 and 4 (P < 0.05) [Table 3].{Table 3}

Discussion

There are various previous studies regarding the analysis of accuracy for panoramic radiographs, [1],[4],[6],[13],[14] conventional spiral computed tomography (MSCT), [1],[2],[3],[4],[5],[15],[16],[17] digital radiography, [18] and cone-beam tomography. [5],[9],[14],[19],[20],[21] However, there are few reports regarding the cone-beam tomography especially with the intraexaminer and interexaminer evaluation [10],[12],[22] and it is only a systematic review published concerning the measurement accuracy and the reliability of landmark identification through CT and CBCT in the maxillofacial area that for the authors more research is required in this field. [8] The results of this study were compared with those in the literature for evaluating the cone-beam reliability and reproducibility for linear measurements.

There was little difference in intraexaminer and interexaminer repeatability. With the limited material available in this study, low-dose CBCT examinations, with radiation dose less than a quarter of conventional values, produced images with a quality sufficient to make reliable linear measurements for the preoperative planning of placement of oral implants. [5]

Interexaminer mean differences were greater than intraexaminer differences [Table 1] and [Table 2]. This can be explained based on the examiner's interpretation of landmark definition and individual anatomic variations. Furthermore, operator experience using CBCT images and software may have influenced the study results by having a greater impact on interexaminer reliability. The other studies reported similar results. [22]

According Lagravere et al, [22] landmarks with variations higher than 1.0 mm would be of clinical significance. The size of the structure being investigated and the magnitude of change to be detected will also influence the clinical significance of landmark identification error. Some measurements in this study had this higher variation, especially interexaminer view.

Kobayashi et al, [20] compared the accuracy of measurement of distance using limited CBCT (LCBCT) and helical CT (HCT). The vertical distance from a reference point to the alveolar ridge was measured in five cadaver mandibles. A significantly smaller measurement error was also observed in their study for LCBCT than for HCT, 1.4% and 2.2%, respectively (P<0.0001). The results of Suomalainen et al, [5] for measurement errors were higher than our study and the study of Kobayashi et al.[20] This can be explained at least in part by the thickness of the cortex of the inferior border of the mandible. Even a minor measurement error leads to a considerable relative difference from the gold standard. In the dentate region, the difficulty of selecting the exact level of the alveolar crest might also explain the higher measure error. [5]

The reconstruction pixel size of 0.25 mm chosen for this study limits accuracy of individual measurements to ± 0.5 mm. Although primary and secondary slice reconstruction thickness will also influence measurement, narrower slices should result in better measurement accuracy. Measurement of wire markers in this study suggest that the i-CAT classic performed well in providing volumes that are relatively free from distortion in vertical and horizontal dimensions. Even though, Ludlow et al, used a grater pixel size (0.5 mm), their error measurements were similar to our present work.

Linear measurement accuracy using the i-CAT measuring tool is comparable to the accuracy reported by other authors. [23] The mean absolute error of 0.13 mm and standard deviation of 0.09 mm in that study were similar with this work [Table 1] and [Table 2].

Cone-beam computed tomography volumes provide the radiologist with the flexibility of choosing the orientation of the reconstructed image layer. The operator has the advantage of constructing the image layer after studying a volume of axial slices. This way, the image layer can be adapted to the anatomy of interest. The measurement errors seen with two-dimensional techniques are reduced with three-dimensional measurement techniques. The small measurement errors seen with two-dimensional measurement techniques in this study suggest that the chosen anatomic points and reference lines had acceptably small deviations from the plane of the panoramic reconstruction. [10]

This study compared the performance of two different examiners for measuring some landmarks from mandibular canal in cone-beam tomography. The use of 2D image as panoramic radiographs for linear measurements may not represent the actual dimensions of the jaw because the magnification and distortion on the panoramic radiographs in different parts of jaws vary due to the changing distance between the rotational center and film and due to the changing rate of movement of the film. In addition, the beam is not aimed horizontally; there is thus some distortion in a vertical direction, [24] being a weakness of this work.

The intraclass correlations between the intraexaminer and interexaminer readings for the different methods showed an almost perfect match. However, the same measurements were evaluated continuously, which may have strengthened the apparent reliability of the measurements. The 1-month interval between the measurements contributes positively to the reliability.

The challenge of this work arises in defining and selecting the points to be measured. This problem is an illustration of the difficulty in using three-dimensional tools to measure anatomic points that have traditionally been defined using two-dimensional image techniques and tools.

While the measured features and measurement techniques differed from previous studies, [6],[7],[8],[9],[10],[11],[12],[13],[14],[15],[16],[17],[18],[19],[20] average measurement errors from 0.2 to 2.1 mm are in line with errors reported for both conventional and cone beam CT. [9],[10],[25]

Until the present moment, in 2007, the only systematic review about the accuracy of measurements and reliability of landmarks identification with computed tomography techniques in the maxillofacial area was published by Lou et al.[8] From the electronic databases search, they encountered eight papers to include in their study. It was concluded that each landmark exhibited a characteristic pattern of error that contributed to measurement inaccuracy, and with repeated practice of landmark identifications, the error can be reduced to within 0.5 mm for 2-D CT. Considerations have to be given to some of the 3-D CT reliability values because they can have diagnosis implications. We agreed with this statement because the measurement error decreases while the length of the slices diminishes.

In conclusion, the measurements demonstrated strong agreement between examiners and significant reliability and reproducibility. Therefore, this methodology can serve as a standard for linear measurement analysis of the topography of mandibular canal and osseous adjacent structures with high accuracy. With the appropriate choice and definition of landmarks, and the availability of measuring tools, this accuracy has the potential of providing unambiguous information for correct diagnosis.

References

1	Ylikontiola L, Moberg K, Huumonen S, Soikkonen K, Oikarinen K. Comparison of three radiographic methods used to locate the mandibular canal in the buccolingual direction before bilateral sagittal split osteotomy. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2002;93:736-42.
2	Gomes AC, Vasconcelos BC, Silva ED, Caldas Ade F Jr, Pita Neto IC. Sensitivity and specificity of pantomography to predict inferior alveolar nerve damage during extraction of impacted lower third molars. J Oral Maxillofac Surg 2008;66:256-9.
3	Valmaseda-Castellon E, Berini-Aytes L, Gay-Escoda C. Inferior alveolar nerve damage after lower third molar surgical extraction: A prospective study of 1117 surgical extractions. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2001;92:377-83.
4	de Melo Albert DG, Gomes AC, do Egito Vasconcelos BC, de Oliveira e Silva ED, Holanda GZ. Comparison of orthopantomographs and conventional tomography images for assessing the relationship between impacted lower third molars and the mandibular canal. J Oral Maxillofac Surg 2006;64:1030-7.
5	Suomalainen A, Vehmas T, Kortesniemi M, Robinson S, Peltola J. Accuracy of linear measurements using dental cone beam and conventional multislice computed tomography. Dentomaxillofac Radiol 2008;37:10-7.
6	Gungor K, Akarslan Z, Akdevelioglu M, Erten H, Semiz M. The precision of the panoramic mandibular index. Dentomaxillofac Radiol 2006;35:442-6.
7	Tyndall DA, Brooks SL. Selection criteria for dental implant site imaging: A position paper of the American Academy of Oral and Maxillofacial radiology. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2000;89:630-7.
8	Lou L, Lagravere MO, Compton S, Major PW, Flores-Mir C. Accuracy of measurements and reliability of landmark identification with computed tomography (CT) techniques in the maxillofacial area: A systematic review. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2007;104:402-11.
9	Lascala CA, Panella J, Marques MM. Analysis of the accuracy of linear measurements obtained by cone beam computed tomography (CBCT-NewTom). Dentomaxillofac Radiol 2004;33:291-4.
10	Ludlow JB, Laster WS, See M, Bailey LJ, Hershey HG. Accuracy of measurements of mandibular anatomy in cone beam computed tomography images. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2007;103:534-42.
11	Friedland B, Donoff B, Dodson TB. The use of 3-dimensional reconstructions to evaluate the anatomic relationship of the mandibular canal and impacted mandibular third molars. J Oral Maxillofac Surg 2008;66:1678-85.
12	Park CH, Abramson ZR, Taba M Jr, Jin Q, Chang J, Kreider JM, et al. Three-dimensional micro-computed tomographic imaging of alveolar bone in experimental bone loss or repair. J Periodontol 2007;78:273-81.
13	Laster WS, Ludlow JB, Bailey LJ, Hershey HG. Accuracy of measurements of mandibular anatomy and prediction of asymmetry in panoramic radiographic images. Dentomaxillofac Radiol 2005;34:343-9.
14	Tantanapornkul W, Okouchi K, Fujiwara Y, Yamashiro M, Maruoka Y, Ohbayashi N, et al. A comparative study of cone-beam computed tomography and conventional panoramic radiography in assessing the topographic relationship between the mandibular canal and impacted third molars. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2007;103:253-9.
15	Valmaseda-Castellon E, Berini-Aytes L, Gay-Escoda C. Supernumerary premolars: Report of 10 cases. Bull Group Int Rech Sci Stomatol Odontol 2001;43:19-25.
16	Cavalcanti MG, Yang J, Ruprecht A, Vannier MW. Accurate linear measurements in the anterior maxilla using orthoradially reformatted spiral computed tomography. Dentomaxillofac Radiol 1999;28:137-40.
17	Todd AD, Gher ME, Quintero G, Richardson AC. Interpretation of linear and computed tomograms in the assessment of implant recipient sites. J Periodontol 1993;64:1243-9.
18	Brito-Junior M, Santos LA, Baleeiro EN, Pego MM, Eleuterio NB, Camilo CC. Linear measurements to determine working length of curved canals with fine files:conventional versus digital radiography. J Oral Sci 2009;51:559-64.
19	Haiter-Neto F, Wenzel A, Gotfredsen E. Diagnostic accuracy of cone beam computed tomography scans compared with intraoral image modalities for detection of caries lesions. Dentomaxillofac Radiol 2008;37:18-22.
20	Kobayashi K, Shimoda S, Nakagawa Y, Yamamoto A. Accuracy in measurement of distance using limited cone-beam computerized tomography. Int J Oral Maxillofac Implants 2004;19:228-31.
21	Neugebauer J, Shirani R, Mischkowski RA, Ritter L, Scheer M, Keeve E, et al. Comparison of cone-beam volumetric imaging and combined plain radiographs for localization of the mandibular canal before removal of impacted lower third molars. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2008;105:633-42;discussion 43.
22	Lagravere MO, Gordon JM, Guedes IH, Flores-Mir C, Carey JP, Heo G, et al. Reliability of traditional cephalometric landmarks as seen in three-dimensional analysis in maxillary expansion treatments. Angle Orthod 2009;79:1047-56.
23	Marmulla R, Wortche R, Muhling J, Hassfeld S. Geometric accuracy of the NewTom 9000 Cone Beam CT. Dentomaxillofac Radiol 2005;34:28- 31.
24	Yang J, Cavalcanti MG, Ruprecht A, Vannier MW. 2-D and 3-D reconstructions of spiral computed tomography in localization of the inferior alveolar canal for dental implants. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1999;87:369-74.
25	Cavalcanti MG, Rocha SS, Vannier MW. Craniofacial measurements based on 3D-CT volume rendering: Implications for clinical applications. Dentomaxillofac Radiol 2004;33:170-6.