|Year : 2022 | Volume
| Issue : 1 | Page : 95-99
Reliability and reproducibility of a conversion factor for grayscale values obtained from CBCTS assessed at various anatomical regions- A retrospective study
TR Prasanna Arvind, Ravindra Kumar Jain
Department of Orthodontics, Saveetha Dental College, Saveetha Institute of Medical and Technical Sciences, Chennai, Tamil Nadu, India
|Date of Submission||04-May-2021|
|Date of Decision||13-Jul-2021|
|Date of Acceptance||29-Sep-2021|
|Date of Web Publication||25-Mar-2022|
Dr. T R Prasanna Arvind
Department of Orthodontics, Saveetha Dental College, Chennai, Tamil Nadu
Source of Support: None, Conflict of Interest: None
| Abstract|| |
Background: Bone density assessment in dental practice is required for dental implant planning and placement, Orthodontic tooth movement, and establishing mid palatal suture maturation. Establishing a linear relationship between grayscale values and Hounsfield units (HUs) in softwares can improve the ease of diagnosis and treatment planning. Aim: To determine a conversion factor and test its reliability and reproducibility for grayscales values to HUs obtained in Cone-beam computed tomography (CBCT) s at various anatomical regions using Galileos Sirona CBCT Viewer Software and Philips DICOM Viewer. Methods: Eighty-seven (87) CBCTs were included in this retrospective study and 343 sites were assessed across different anatomical regions. Sites assessed were isolated in both softwares at identical locations. Grids were used for standardizing reference planes at different anatomical sites. Reference landmarks were isolated 1) between anterior interdental regions 2) between posterior interdental regions, and 3) at radio-opaque regions. Grayscale value was divided with HU to obtain conversion factors at different sites, which was the primary outcome of the study. The reproducibility of the factor obtained was also assessed to improve its clinical correlation. Results: In anterior interdental regions, grayscale values were 7.6 times greater than HUs. In posterior interdental regions, gray scale values were 4.5 times greater than HUs. The reliability of values obtained was verified using Kappa's correlation test. In radio-opaque regions, grayscale values were 1.4 times greater than HUs and highly reliable (r = 0.972). Conclusion: In well-exposed regions, a defined conversion factor can be established between grayscale values and HUs in Galileos Software. This factor determined is highly reproducible and reliable in radio-opaque regions, and adequately reliable in anterior and posterior interdental regions.
Keywords: Cone-beam computed tomography, Hounsfield units, grayscale values
|How to cite this article:|
Prasanna Arvind T R, Jain RK. Reliability and reproducibility of a conversion factor for grayscale values obtained from CBCTS assessed at various anatomical regions- A retrospective study. J Indian Acad Oral Med Radiol 2022;34:95-9
|How to cite this URL:|
Prasanna Arvind T R, Jain RK. Reliability and reproducibility of a conversion factor for grayscale values obtained from CBCTS assessed at various anatomical regions- A retrospective study. J Indian Acad Oral Med Radiol [serial online] 2022 [cited 2022 May 27];34:95-9. Available from: https://www.jiaomr.in/text.asp?2022/34/1/95/340727
| Introduction|| |
Computed tomography (CT) images are used for the evaluation of soft and hard tissues aiding in the diagnosis of pathologic and traumatic lesions in the craniofacial region. CT has a standard design to measure beam attenuation by the body tissues, which is referred to as Hounsfield Unit (HU). HU is used to describe and characterize bone density permitting a more intricate assessment of the alveolar region and its adaptive environment. However, CT scans are invasive and carry with them it the additional risk of exposing the individual to high levels of radiation that can produce stochastic effects. Fears of increased and unnecessary exposure to the entire maxillofacial region have led to the increased usage of cone-beam computed tomography (CBCT) scans for evaluating bone density.
CBCT scans evaluate bone density using grayscale levels, which is different from conventional HUs. Both units utilize beam attenuation levels to analyze bone density, however, the accuracy of grayscale values to HUs is questionable. It should be taken into account that the gray level is not the same as true HU. The HU scale used in CT takes into account the linear attenuation coefficient of water. In CBCT, hard tissues absorb greater amounts of X-ray energy compared with soft tissues creating different shades of gray on 3-D images. Thus, CBCT uses different gray values to define morphology and measure density., Various investigators have proposed machine-specific methods for converting values obtained with CBCT systems into HU values. With one conversion method, meticulous measurement and linear regression analysis were needed to convert CBCT numbers into HU values., Although average values for different materials were mostly predictable, HU values obtained were excessively variable, making them unsuitable for clinical use. For CBCT systems, the above factors present many challenges in transforming CT numbers into HUs that accurately represent the physical densities of all points within the scanned object.,
The current study deals with determining a conversion factor between the grayscale values obtained in Galileos Software and HUs projected in Philips DICOM viewer. This would enable us to establish standard values at desired regions of interest, optimizing imaging diagnosis. It seeks to establish a connection between the grayscale values that are highly arbitrary and standard correlation coefficients. Do the grayscale values obtained in Galileos Viewer possess a correlation coefficient between different regions that can provide for easier reference values during diagnosis? Considering the ever-increasing clinical use of the gray level and the advantages of CBCT over CT, the present study was undertaken to establish a conversion factor and test its reliability and reproducibility in various anatomical regions using Galileos Sirona CBCT Viewer Software and Philips DICOM Viewer.
| Materials and Methods|| |
Eighty-seven (87) CBCTs included in the retrospective study were obtained from the Department of Oral Radiology. Sample size calculation was determined from a similar study conducted. About 77 samples were required in each group to obtain a power of 90% with an effect size of 0.2 and alpha-error of 0.5. Adult patients who reported to dental OPD and required CBCT imaging were included in the study. The exclusion criteria were scans of patients with trauma/pathologic injuries of alveolar bone, systemic conditions affecting bone density, and related medications. About 87 CBCTs were included in the current study after eliminating scans with poor diagnostic quality. All procedures performed in this study were following the Ethical clearance obtained from the Institutional Review Board (IHEC/SDC/ORTHO-1803/21/37) and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
CBCTs obtained were analyzed using Galileos Software from Dentsply Sirona. They were converted into DICOM format for usage in Philips DICOM viewer that projects grayscale values as HUs. CBCT scan settings were set at 120 kV, 6 mA, 18,817 mS, FOV 80 mm, and slice thickness of 0.1 mm.
Philips DICOM Viewer is a CT viewer software that allows bone quality to be assessed in HUs due to its in-built software that allows for a recalibration of grayscale values. Site standardization was done using customized grids available in Philips DICOM viewer that can allow for easier reconstruction of the images used in the study. Identical sites were maintained in both the CBCT viewer and CT viewer, and these sites were standardized [Figure 1].
|Figure 1: Standardization of anatomical sites assessed in both Galileos CBCT Viewer and Philips DICOM Viewer|
Click here to view
In-built grids in Philips DICOM software were utilized for analyzing the region of preference. The grids were used to segregate the areas and bone density values were analyzed independently in each software [Figure 2]. Grayscale values were obtained in Galileos Software and Hounsfield units were obtained in Philips DICOM software.
|Figure 2: In-built customized grids used to standardize sites across various regions in Philips DICOM Viewer|
Click here to view
The different sites assessed were divided into anterior and posterior regions. The region between the centrals, laterals, laterals, and canines, canines and first premolars were classified as the anterior regions. The posterior regions were classified as the region between the first and second premolars, second premolar and first molar, first and second molars [Figure 3]. Radio-opaque regions (infra zygomatic crest region, bone thickness) were assessed independently.
|Figure 3: Various anatomical regions assessed in the study identified using ellipses catering to the region of interest (ROI)|
Click here to view
Grayscale values projected in Galileos CBCT Viewer were divided by the corresponding HUs obtained in Philips DICOM Viewer to obtain a conversion factor relevant to the anatomical site. Reproducibility of the values obtained was verified using Kappa's correlation test to determine the accuracy of the values at a particular anatomical site.
| Results|| |
[Table 1] corresponds to the mean and standard deviation of grayscale values and HUs obtained in different anatomical regions. The grayscale value obtained at the interdental regions of anterior teeth was 2188 ± 315.4 and the corresponding HU was 319 ± 116.9. The conversion factor obtained was 1: 7.6. The grayscale value obtained at the interdental regions of posterior teeth was 3379 ± 612.8 and the corresponding HU was 742 ± 202.4. The conversion factor obtained was 1: 4.5. The grayscale value obtained at radio-opaque regions was 1962 ± 387.8 and the corresponding HU was 1387 ± 319.6. The conversion factor obtained was 1: 1.4.
|Table 1: Mean and standard deviation of gray scale values and Hounsfield units obtained in different anatomical regions|
Click here to view
Kappa correlation test was done to verify the reproducibility of the values obtained at the respective sites [Table 2]. Thirty (30) CBCTs were re-assessed by the principal investigator to assess validity. The anterior and posterior interdental regions showed adequate reproducibility and operator validity with correlation coefficients of 0.786 and 0.814, respectively. Radio-opaque regions showed a high degree of correlation (r = 0.972) and were highly reproducible.
|Table 2: Kappa intraclass correlation coefficient for 30 CBCTs re-evaluated by the same operator after individual site assessment to test reproducibility of values|
Click here to view
| Discussion|| |
CBCT scans have shown tremendous growth and rapid development in the field of maxillofacial imaging over the past two decades. It is convenient, reliable, and less invasive in terms of radiation when compared to conventional CT scans. Moreover, diagnostic details that can be cleaned from tomographic scans are limited in dentistry and can be satisfactorily carried out by CBCT scans. Multiple studies have shown a high correlation among values obtained for bone density using CBCT machines to HUs. These studies have emphasized determining a linear coefficient between the grayscale values obtained that can be used in diagnosis and treatment planning., In the present study, we have utilized grayscale values obtained from Sirona Dental software and obtained the values in HUs in Philips DICOM viewer.
A significant finding obtained from the present study is the high reliability of the correlation factor obtained in the highly radio-opaque regions. The infra zygomatic crest (IZC) region is highly radio-opaque and presents in the posterior region of the maxilla. Considering the high reproducibility of the values obtained, we can conclude that IZC site assessment can be carried out satisfactorily using CBCT scans rather than going for invasive CT scans. While the exact magnitude of the bone density might vary between units, a conversion factor obtained might make clinical evaluation much easier. This finding has significant value in the usage of CBCTs for establishing bone density and bone thickness values. These regions show a lesser degree of beam attenuation and can be reliably analyzed using CBCT scans. Moreover, implant placement accuracy, site assessment of implant placement all can be predicted to a great degree of accuracy. This might support the claims of studies that have utilized CBCTs to study bone density values since they are highly reproducible by type.,,
Valiyaparambil et al. evaluated hard-tissue equivalent materials; the results of the present study showed a weaker linear correlation in terms of the numeric values compared to the present study. However, both studies indicated a strong linear correlation between HU and gray level and the differences in the results of these two studies might be due to the differences in sample collection.
In a study by Reeves et al., comparative evaluations were carried out between CBCT systems, with the use of artificial materials placed in the patients' oral cavities. However, comparisons were not carried out with the true HU of the imaging system, and the HU values were presented as values achieved through the relationship acquired from previous in vitro studies.
Currently, CBCT manufacturers provide gray levels, which are not actual Hounsfield units, which make it difficult to assess bone quality from a CBCT data set. A method that converts gray levels taken from CBCT data sets into HUs would standardize and allow the comparison of bone quality from machine to machine within a small range. This capability along with the decreased patient radiation exposure, ease of access, a greater resolution than medical CT, and affordability should render CBCT as the imaging modality of choice to assess alveolar bone for dental implant placement and to measure bone density.
| Limitations And Future Prospects|| |
Establishing conversion factors for different anatomical regions can be used to standardize image acquisition techniques. Moreover, it can also help us isolate areas that are well exposed and under-exposed. Doing so would allow us to practice greater pragmatism at different sites of mini-implant placement. While a direct clinical application of this study might not be readily feasible, we could potentially translate these findings during implant placement to identify regions that are well supported by adequate bone density. The findings obtained from the current study apply only to the Galileos Sirona Dental CBCT scanner. These findings cannot be transferred to any other CBCT unit using different exposure protocols. It does not seek to establish whether grayscales obtained in CBCT units might replace HUs. Thus, there is a need for a conversion factor to use a similar methodology for other CBCT machines. In this study, this conversion factor was determined for the Galileos Sirona CBCT Viewer; thus its use is applicable only for this particular machine. Conversion factors for other machines can be determined using similar methodologies like as the one used in this study.
| Conclusions|| |
The conversion factors for grayscale values to HUs determined in the anterior and posterior interdental regions are highly reliable and reproducible (r = 0.786 and 0.814, respectively).
Conversion factor for grayscale values to HUs determined at radio-opaque regions is highly accurate and reproducible (r = 0.972).
I would like to acknowledge the efforts of my co-authors along with my Department Head for encouraging us to get involved in interdisciplinary research topics involving matters of concern for orthodontic purposes.
Declaration of patient consent
The authors certify that they have obtained all appropriate patient consent forms. In the form, the patient(s) has/have given his/her/their consent for his/her/their images and other clinical information to be reported in the journal. The patients understand that their names and initials will not be published and due efforts will be made to conceal their identity, but anonymity cannot be guaranteed.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Shokri A, Ghanbari M, Maleki FH, Ramezani L, Amini P, Tapak L. Relationship of gray values in cone beam computed tomography and bone mineral density obtained by dual energy X-ray absorptiometry. Oral Surg Oral Med Oral Pathol Oral Radiol 2019;128:319-31.
Mah P, Reeves TE, McDavid WD. Deriving Hounsfield units using grey levels in cone beam computed tomography. Dentomaxillofac Radiol 2010;39:323-35.
Kim TH, Lee DY, Jung SK. Comparison of trabecular bone mineral density measurement using Hounsfield unit and trabecular microstructure in orthodontic patients using cone-beam computed tomography. Appl Sci 2021;11:1028.
Nomura Y, Watanabe H, Honda E, Kurabayashi T. Reliability of voxel values from cone-beam computed tomography for dental use in evaluating bone mineral density. Clin Oral Implants Res 2010;21:558-62.
Reeves TE, Mah P, McDavid WD. Deriving Hounsfield units using grey levels in cone beam CT: A clinical application. Dentomaxillofac Radiol 2012;41:500-8.
Kamaruddin N, Rajion ZA, Yusof A, Aziz ME. Relationship Between Hounsfield Unit in CT Scan and Gray Scale in CBCT. AIP Conference Proceedings. Vol 1791. AIP Publishing LLC; 2016. p. 020005.
Abe T, Tateoka K, Saito Y, Nakazawa T, Yano M, Nakata K, et al
. Method for converting cone-beam CT values into Hounsfield units for radiation treatment planning. Int J Med Phys Clin Eng Radiat Oncol 2017;6:361-75.
Bryant J. Deriving Hounsfield units from the grey scale of a CBCT? Dentomaxillofac Radiol 2011;40:65.
Badey A, Bodez V, Khamphan C, Jaegle E, Alayrach ME, Martinez P, et al
. 31 Evaluation of Hounsfield Unit correction method on Cone-Beam CT for dose calculation strategies. Physica Medica: European Journal of Medical Physics. 2018;56:19.
Buenger F, Eckardt N, Sakr Y, Senft C, Schwarz F. Correlation of bone density values of quantitative computed tomography and Hounsfield units measured in native computed tomography in 902 vertebral bodies. World Neurosurg 2021;151:e599-606.
Ludlow JB, Ivanovic M. Comparative dosimetry of dental CBCT devices and 64-slice CT for oral and maxillofacial radiology. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2008;106:106-14.
Razi T, Emamverdizadeh P, Nilavar N, Razi S. Comparison of the Hounsfield unit in CT scan with the gray level in cone-beam CT. J Dent Res Dent Clin Dent Prospects 2019;13:177-82.
Cassetta M, Stefanelli LV, Pacifici A, Pacifici L, Barbato E. How accurate is CBCT in measuring bone density? A comparative CBCT-CT in vitro study. Clin Implant Dent Relat Res 2014;16:471-8.
Zhang J, Tian Y, Liu Y, Liu Q, Zhang J, Liu J, et al
. Alveolar bone mineral density measurement using CBCT images. Neurosci Biomed Eng 2017;5:44-9.
Kim DG. Can dental cone beam computed tomography assess bone mineral density? J Bone Metabol 2014;21:117-26.
Chugh T, Ganeshkar SV, Revankar AV, Jain AK. Quantitative assessment of interradicular bone density in the maxilla and mandible: implications in clinical orthodontics. Progress in orthodontics. 2013;14:1-8.
Haristoy RA, Valiyaparambil JV, Mallya SM. Correlation of CBCT gray scale values with bone densities. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2009;4:e28.
[Figure 1], [Figure 2], [Figure 3]
[Table 1], [Table 2]