A simplified low‐cost phantom for image quality assessment of dental cone beam computed tomography unit

Abstract

Introduction

A standardised testing protocol for evaluation of a wide range of dental cone beam computed tomography (CBCT) performance and image quality (IQ) parameters is still limited and commercially available testing tool is unaffordable by some centres. This study aims to assess the performance of a low‐cost fabricated phantom for image quality assessment (IQA) of digital CBCT unit.

Methods

A customised polymethyl methacrylate (PMMA) cylindrical phantom was developed for performance evaluation of Planmeca ProMax 3D Mid digital dental CBCT unit. The fabricated phantom consists of four different layers for testing specific IQ parameters such as CT number accuracy and uniformity, noise and CT number linearity. The phantom was scanned using common scanning protocols in clinical routine (90.0 kV, 8.0 mA and 13.6 s). In region‐of‐interest (ROI) analysis, the mean CT numbers (in Hounsfield unit, HU) and noise for water and air were determined and compared with the reference values (0 HU for water and −1000 HU for air). For linearity test, the correlation between the measured HU of different inserts with their density was studied.

Results

The average CT number were −994.1 HU and −2.4 HU, for air and water, respectively and the differences were within the recommended acceptable limit. The linearity test showed a strong positive correlation (R ² = 0.9693) between the measured HU and their densities.

Conclusion

The fabricated IQ phantom serves as a simple and affordable testing tool for digital dental CBCT imaging.

This study highlights the designation and fabrication of a simple and affordable phantom for quality assurance (QA) test suitable for digital dental CBCT system.

Introduction

The growing demands of advanced technologies in dental imaging have led to introduction of digital cone‐beam computed tomography (CBCT) technique, which offers faster image acquisition with improved image quality. ¹ , ² Since its introduction, CBCT has become a well‐established and important diagnostic tool in maxillofacial, head and neck imaging. ³ , ⁴ , ⁵ The main advantages of CBCT technology are the ability to view and analyse the three‐dimensional (3D) images of dental structures, soft tissues, nerve paths and bone within the craniofacial region and overcome the tissue magnification. Besides, it is preferable as the superimposition of structures found in 2D images such as orthopantomography (OPG) or panoramic imaging can be eliminated. ⁶ , ⁷ , ⁸

Since the tomographic acquisition is employed for reconstruction of the 3D images in dental CBCT, thus the radiation dose from CBCT is higher than other conventional dental imaging modalities. ⁹ , ¹⁰ , ¹¹ In diagnostic radiology, the major concern is to ensure the use of radiation exposure is justified and kept as low as reasonably achievable (ALARA) level while maintaining optimum diagnostic quality, as recommended by the International Commission on Radiological Protection (ICRP). ⁸ , ¹² , ¹³ American Academy of Oral and Maxillofacial Radiology recommends that a quality assurance (QA) program should emphasis on the analysis of image quality (IQ) and dosimetry QA parameters. ⁵ , ⁸ , ¹³ , ¹⁴ , ¹⁵ The SEDENTEXCT project initiated by European Commission (EC) has established a guideline on dental QA program, which comprises the development of a suitable QA test tool and the justification of exposure protocols in order to optimise patient safety. ¹⁶ , ¹⁷ , ¹⁸

Outcomes from routine QA testing should be compared with the established baseline by the regulatory bodies to ensure optimum IQ performance with minimal radiation dose received by patients. ¹⁷ Current QA program for dental CBCT system mainly focuses on the performance parameters of the panoramic and cephalometric views and does not encompass the wide range of performance parameters including tomographic view (CT part). Another limitation in performing routine QA test for dental CBCT is unaffordable testing tool as most of the commercially available phantom in the market are expensive. ¹⁹ , ²⁰ Besides, previous literatures reported that most of the phantoms do not allow wide range of IQ parameters to be evaluated as suggested by European Commission. ¹⁹ , ²⁰ Thus, an in‐house phantom is demanded for certain centres where proper test tool may not be readily available and unaffordable.

This study aims to assess a low‐cost phantom specifically designed and fabricated for image quality assurance (IQA) evaluation of the digital dental CBCT system. A wide range of IQ performance parameters of the digital dental CBCT system were evaluated using the fabricated phantom. The evaluation focused on the IQ parameters for the tomographic view or CT part of CBCT that covers CT number accuracy, consistency, uniformity, noise and CT number linearity.

Methods

Phantom fabrication

A customised phantom was developed using polymethyl methacrylate (PMMA) with a head‐sized (approximately 10 cm diameter) to simulate the attenuation and radiographic properties of human soft tissue in an adult body head size (Fig. 1). The design of the phantom was based on the features and parameters of the previous phantoms from literatures. ³ , ¹⁹ , ²⁰ All fabrication and development works were performed at the Physics Engineering Workshop, Universiti Sains Malaysia (USM), Malaysia. A custom phantom holder with circular shape (15 cm in diameter) was also fabricated as base or support to hold the phantom on the digital dental CBCT unit during phantom scanning. It is positioned at the bottom of phantom as shown in Figure 1 (left). The phantom consists of four layers specifically for testing various parameter of the image quality (IQ) tests. The first top 3 layers are used to evaluate CT number accuracy and uniformity and noise within different uniform background regions (water, PMMA and air).

Figure 1.

The fabricated phantom with different layers and phantom holder at the bottom (left) and the cylindrical insert for CT linearity test (right).

The first and second layers consist of hollow cylindrical layer constructed from a PMMA tube with 10 cm diameter and 1 cm wall thickness. During phantom scanning, the first layer was filled with distilled water, while the second layer remained empty to facilitate air measurements. The third and fourth layers are constructed with solid PMMA material with cylindrical shape. The fourth layer was drilled to form 5 hollow cylinders or cavity to accommodate for different inserts (3 cm in diameter). Each cavity is inserted with different inserts made of various materials (air, nylon, brass and Teflon), used for CT number linearity test (Fig. 1). The insert materials within the fourth layer of the phantom simulate different tissues within the maxillofacial region and these features applicable for the dental radiography purposes. The brass insert simulates the dental fillings (amalgam), Teflon or PTFE insert simulates the dense bone (jawbone, enamel and dentin), nylon and PMMA inserts simulate the soft tissues (pulp and gum), while the air region simulates air cavities inside the maxillofacial region of human. ²¹

Phantom study

The phantom scanning was performed using the Planmeca ProMax 3D Mid CBCT unit (Planmeca Oy, Helsinki, Finland), a digital dental CBCT system at Imaging Unit, Advanced Medical and Dental Institute, Universiti Sains Malaysia (USM). During the phantom scanning, the customised phantom was positioned on the phantom holder (at the isocentre of the gantry), which then inserted into adapter holes on the patient support table. The phantom scanning was acquired at exposure setting of 90.0 kV, 8.0 mA and 13.6 s with field of view (FOV) of medium (M) patient size. It is the common scanning protocol used in local clinical procedures. The slice thickness selected was 1.2 mm for all phantom images. The scanning was repeated for 3–4 times for each different phantom layers using constant exposure setting and slice thickness.

Image quality performance test

The acquired images of different phantom layers were then displayed on the CBCT workstation. An associated CBCT software, Planmeca Romexis dental software (Version 3.6.0.r) (Helsinki, Finland), was used for IQ analysis. Few IQ parameters were evaluated such as CT number accuracy, consistency and noise (SD). The region of interest (ROI) analysis was employed to determine the attenuation value represented as CT number (in Hounsfield unit, HU) and noise (in standard deviation, SD) of the reference materials such as water, air and PMMA (homogeneous background on the first to third layers).

For CT number accuracy test, a circular ROI of approximately 200 mm² size was drawn at the centre of phantom images, as shown in Figure 2 (left). The average HU and SD values were recorded for each repeated scanning. The deviation values were determined (CT_measured – CT_standard) and the acceptable levels are ±5 HU and ±10 HU for water and air, as recommended by Malaysian Ministry of Health (MOH). ²² For CT number uniformity, five equal sized (approximately 200 mm²) of circular ROIs were placed at central and four at peripheral part of the phantom image, as shown in Figure 2 (right). The drawn ROIs were copied from the first measurement to maintain similar ROIs size and to ensure accurate attenuation measurement.

Figure 2.

The position of the drawn ROIs for the calculation of CT number accuracy (left) and CT number uniformity (right).

Data analysis

CT number uniformity is determined by calculating the maximum deviation of the measured attenuation (mean HU) values of each drawn ROIs on the same image slice. The maximum deviation is calculated by the equation (max HU_avg – min HU_avg). ¹³ , ¹⁸ , ²³ Based on the recommendation by Malaysian MOH, the maximum deviation for head phantom should be within ≤ ±5 HU. ²² The comparative study was performed by comparing the measured CT HU and noise with the acceptable limit to determine the system performance level. For CT number linearity, the Pearson correlation coefficient was determined to investigate the relationship (strength) between the measured HU with the materials density. The correlations were also plotted to evaluate the significant linear correlation between the paired measurement parameters.

Results

CT number accuracy and noise

The Romexis software generates the mean HU value and SD within the drawn ROIs on the axial phantom images. From the findings, the range of HU for water is between −110 and +123 HU. Based on the manufacturer's (Planmeca) requirements, the standard HU range for water is between −150 and +150 HU, and the CT number test should be performed on daily basis. ²² Table 1 shows the measured CT number (in HU) for different reference materials of the phantom such as water, PMMA and air. The mean values of measured CT number for both air and water within the phantom layers were −994.1 HU and −2.4 HU, respectively. Theoretically, the standard CT number for water is 0 HU and air is −1000 HU.

Table 1.

The measured attenuation (CT number) and the noise (or SD) values for different reference materials within the phantom.

Reference materials	Measured readings		Deviation (HU) 1
	Average CT no. (HU)	Average SD, Noise (HU)
Water	−2.38	27.49	−2.38
PMMA	−3.60	27.38	−3.60
Air	−994.09	11.76	−5.91

¹ The deviation is calculated by (CT_measured – CT_standard) values.

Both water and air are used as reference materials in standard measurement of CT number. The deviation between the measured CT values with the reference or standard value were −5.91 HU and −2.38 HU for air and water, respectively (as shown in right column in Table 1). The deviation values were within the acceptable level of ±5 HU and ± 10 HU for water and air, respectively, as recommended by Malaysian MOH. ²²

CT number consistency and uniformity

For CT number consistency, the ROIs measurements were repeated on different axial slices of the same scanned phantom. While for CT number uniformity, the measurements were repeated on the same axial slices, but at different ROIs locations. From the findings, it can be observed also that the repeated measurement of HU ± SD values on different axial slices of phantom images of the same scanning were constant and within the acceptable limit of ±10 HU for air and ±5 HU and water, as recommended by Malaysian MOH. ²²

Theoretically, scanning a uniform phantom background (water or air‐filled phantom) using similar exposure parameters setting should produce the similar pixel value (HU ± SD) and similar amount of noise. It shows that if the same phantom is scanned with similar exposure setting, the CT number of the reconstructed phantom image of similar reconstruction algorithm should be consistent and not be affected by other factors. Ideally, CT images with uniform background are expected to show consistent attenuation (with acceptable noise index) across the image. ²⁴ Therefore, evaluating the uniformity of CT number is essential to ensure CBCT scanner's capable to produce a homogeneous pixel value (HU) with small variations as compared to mean pixel values.

CT number linearity

Attenuation value or CT number increases with an increase in density and effective atomic number of the insert materials. Therefore, the attenuations or CT numbers are dependent on density of the tissues or material. Figure 3 (left) shows the ROIs analysis used to determine the phantom inserts that consists of various materials such as air (1), PMMA (2), nylon (3), Teflon (4) and brass (as shown in Fig. 3‐left). These selected tissue‐equivalent materials mimic the density range of tissues within the maxillofacial region that is likely to be investigated in clinical CBCT imaging such as teeth and soft tissues.

Figure 3.

The measurement of CT number of various materials using ROIs (left) and the graph of CT number linearity (right).

The lowest CT number measured was observed at air region, while the highest CT values are Teflon and brass, as they are the densest materials. For CT number linearity test, the average measured CT number (or HU) for all the inserts were plotted against their electron density. Figure 3 shows the relationship between the HU and the electron density of the insert material. From the linearity plot, a strong positive correlation (R ² = 0.9693) was observed between the measured HU and their densities, and this shows that the system is performing at optimal level and the findings of CT HU linearity produced by the CBCT unit is acceptable.

Discussions

From the results, it can be observed that the measured CT number of PMMA is higher (1.2 HU) as compared to water. This is because the physical density of PMMA is higher as compared to water. However, the measured CT number of PMMA is still within the acceptable limit by the regulatory body. For image noise, the average SD values for both water and PMMA are comparable. However, the noise is lower in air as compared to water and PMMA. There are few parameters that affect the CT number and noise such as beam hardening, scatter and exposure parameters, which should be maintained at constant setting during the phantom scanning. The accuracy of the CT number should always be checked because it is important in diagnosis as the radiologist will used the value of CT numbers to differentiate pathologies from healthy tissues.

This work has demonstrated that the proposed in‐house phantom is suitable for evaluating wide range of IQ performance parameters specifically for tomographic or CT part of the digital dental CBCT system. The cost of phantom fabrication is much cheaper (development cost around 216 USD) and affordable as compared to the current commercial CBCT phantom in the market (with approximate price range between 1.5 K and 6.5 K USD). Although the commercial QA phantom for CT scanner such as CATPHAN is recommended as an alternative for CBCT performance testing; however, it is bulky, and its large size cannot fit the limited FOV of CBCT imaging. ¹⁹ As reported by previous study, CT number fluctuations are noticeable in larger phantom with smaller FOV due to truncation effect and increase in scattered radiation. ²⁴ Thus, the phantom size should be designed to be fully covered within the FOV. This approach is particularly useful when appropriate tool is not accessible, or the commercial phantoms are not readily available due to expensive price.

Conclusion

The developed dental IQ phantom has been evaluated and can be considered as an alternative tool for QA testing that cover several IQ parameters specifically for CT part (tomographic) of the dental CBCT system such as CT number accuracy and consistency, image noise and CT number linearity. The fabricated phantom also serves as an affordable phantom for medical centre which unable to afford a commercialised phantom for their local routine QA testing program. It can also be concluded that our digital dental system, the Planmeca ProMax 3D Mid CBCT unit (Planmeca, Helsinki, Finland), was in optimum performance and has passed the standard IQ performance test.

Conflict of Interest

The authors declare no conflict of interest.

Data Availability Statement

The data that supports the findings of the study are available on request to the corresponding author.