|
 
 |
| ORIGINAL ARTICLE |
|
| Year : 2020 | Volume
: 17
| Issue : 1 | Page : 21-25 |
|
Evaluation of the computed tomography number for water, field uniformity, image noise and contrast resolutions in Kano Metropolis, Nigeria
Mohammed Sidi, Usman Yahaya Hussain, Anas Ya'u, Idris Garba
Department of Medical Radiography, Faculty of Allied Health Sciences, Bayero University, Kano, Nigeria
| Date of Submission | 18-Feb-2020 |
| Date of Acceptance | 10-Mar-2020 |
| Date of Web Publication | 30-May-2020 |
Correspondence Address:
Mr. Mohammed Sidi
Department of Medical Radiography, Faculty of Allied Health Sciences, Bayero University, Kano
Nigeria
 Source of Support: None, Conflict of Interest: None  | Check |
DOI: 10.4103/njbcs.njbcs_10_20
Context: Quality control (QC) tests for computed tomography (CT) scanners are primarily concerned with the maintenance of CT scanner at the optimum operational conditions for providing the required diagnostic information at the least possible exposure to ionising radiation. Aims: The study was aimed at evaluating a CT number for water, field uniformity, image noise and contrast resolutions in Kano metropolis using phantom and assuring optimum operational condition of the equipment. Materials and Methods: A prospective, crosssectional design was employed. Four scanners tagged CT scanner I–IV were included in the study. CT number for water, field uniformity, noise, as well as highcontrast resolution (HCR) and lowcontrast resolution were tested using head CT water phantom. Results: CT scanner I, III and IV have passed a CT number for water and field uniformity tests (0.3, −0.48 and − 1.36), but scanner II failed the test (−4.38). However, the entire scanners failed noise (4.5, 8.75, 4.90 and 4.93); while only scanner III passed the HCR QC test (bars were countable at the 3rd bar pattern), scanners I, II and IV failed HCR test (bars were countable at 5th, 6th and 4th bar patterns). All the studied scanners failed lowcontrast resolution test (margins appear sharp at the 4th hole) in scanners I, II and IV and the hole was not demonstrated in the phantom image of scanner III. Conclusion: Majority of the CT scanners in Kano metropolis passed the tests for the CT number for water and field uniformity but revealed some inadequate levels of noise, HCR and lowcontrast resolution tests.
Keywords: Computed tomography number for water, image noise, Kano metropolis, low-contrast resolution
How to cite this article:
Sidi M, Hussain UY, Ya'u A, Garba I. Evaluation of the computed tomography number for water, field uniformity, image noise and contrast resolutions in Kano Metropolis, Nigeria. Niger J Basic Clin Sci 2020;17:21-5 |
How to cite this URL:
Sidi M, Hussain UY, Ya'u A, Garba I. Evaluation of the computed tomography number for water, field uniformity, image noise and contrast resolutions in Kano Metropolis, Nigeria. Niger J Basic Clin Sci [serial online] 2020 [cited 2023 Jun 10];17:21-5. Available from: https://www.njbcs.net/text.asp?2020/17/1/21/285460 |
| Introduction | |  |
Computed tomography (CT) scanners create cross-sectional images of the human body with a high radiographic contrast for diagnosis of pathological conditions.[1] It was introduced into clinical practice in 1972 as head only scanner, however, due to rapid technological advancement of the scanner and image reconstruction techniques; modern systems are now capable of performing whole-body examination within a couple of seconds for oncologic staging and follow-up or for trauma imaging, other examinations such as cardiac imaging, virtual colonoscopy, gout imaging and whole-organ perfusion imaging have widened the application profile of CT.[2],[3] High radiation dose is the major drawback of CT when compared with conventional radiography.[4] An average effective dose received by a patient during one-head CT examinations is equivalent to an effective dose of 100 posterior anterior chest X-rays.[5] The complexity of this technology continues to increase as does its potential to deliver substantial doses to patients, consequently; the need for routine quality control (QC) tests. QC tests for CT scanners are primarily concerned with the maintenance of the scanner at the optimum operational condition for providing the required diagnostic information at the least possible exposure to ionising radiation.[6] Routine QC ensures that the equipment is operating appropriately so that the dose is optimised for the necessary image quality.[7] Quality of CT images can be assessed by qualitative and quantitative measurements of physical parameters such as CT number for water, field uniformity, noise, high-contrast resolution (HCR) and low-contrast resolution using suitable test phantom.[3] The CT number for water, field uniformity and image noise are daily QC tests that are within the limits of radiographer.[4],[7] Meanwhile, HCR and low-contrast resolution tests are done on post-instillation and on monthly basis.[5],[7]
The CT numbers for water is defined as the density assigned to a voxel CT scan on an arbitrary scale on which the value is set at zero.[8] Test for CT number is done to assure the equipment manufacturer specifications for CT number. Range of 0 ± 3 at the centre of image and 0 ± 5 Hounsfield unit (HU) at the peripheral locations is acceptable. Possible causes for the CT number of water to be out of range are mis-calibration of the algorithm generating CT numbers.[5] Field uniformity refers to CT number (HU) variations in a uniform field (usually a water or water-equivalent phantom), and it is assessed by comparing the attenuation in a region of interest (ROI) at the centre of the uniform field versus along the edges.[9] Its measurements are important to ensure that cupping and beam hardening artefacts are avoided. The standard for field uniformity centre ROI measured < 3 HU and the four ROIs in the periphery measured within the <±5 HU of centre measurement are acceptable.[5] Image noise is a local statistical fluctuation in CT numbers of individual picture elements of a homogenous ROI. The magnitude of noise is indicated as the standard deviation (SD) of the HUs in a defined ROI in a uniform area of the image.[3] Noise in CT is mainly related to the number of detected photons, matrix size (pixel size), slice thickness, algorithm, electronic noise (detector electronics), scattered radiation and object size. It limits low-contrast resolution and may hide anatomy similar to surrounding tissue. The SD for noise should be ±3. Since CT numbers range from ± 1000 HU, the noise is <0.3%. The maximum SD between the centre of ROI and any peripheral ROI is <±5 HU.[5] HCR is the ability to differentiate two high-contrast objects placed close together. Its measurements are performed with objects which have a high contrast (contrast difference of 12% or greater) from uniform background. The SD for an ROI within the normal bar pattern should equal 37 ± 4 for standard algorithm.[10] This test should be performed upon acceptance of equipment and monthly.[5] Low-contrast resolution is the ability to differentiate objects with slightly different densities. Its test measures scanners ability to detect objects that vary only slightly from its background [11]. The basis of this test is that the number of holes visualised on the phantom image is determined, and the smallest hole that should be visualised is the central or third hole (5 mm in diameter). This is especially important when trying to detect low-density tumors that lie in soft tissues. CT has excellent low-contrast detectability as it can resolve density differences as little as 0.1%. Arguably, this is perhaps the most important QC performance test.[5] The test is completed upon installation and then performed biannually thereafter.[10] The findings of the study are expected to serve as a baseline both for making recommendations to the relevant authorities to be performing QC tests where it is not being performed and for corrective measures where the scanner failed the test. It will also serve as a guide to the radiographers and radiologists. The study aims at evaluating some image quality parameters of CT scanners in Kano metropolis using head phantom.
| Materials and Methods | | |
A prospective, cross-sectional design was employed in the study. The study was conducted from 1st April 2019 to 30th November 2019. Ethical approval was sought and obtained from the Human Research and Ethics Committee of Aminu Kano Teaching Hospital, Kano. All the centres in the Kano metropolis with functional CT scanners and which agreed to participate were included in the study. There were eight installed CT scanners in the study area, but only four gave consent for the work. These are tagged CT scanner I–IV. Two centres had CT downtime as at the time of the study, while the remaining two facilities declined to participate. CT scanner I is a 160 slice Toshiba Aquilion (TSX-303A), which was manufactured in 2011 and installed in 2015. CT scanner II is a 4 slice General Electronics BrightSpeed CT (5234959-30), which was manufactured in 2007 and installed in 2009, while CT scanners III is 16 slice GE scanner Brivo 515, which was manufactured in 2013 and installed in 2018, and CT scanner IV is a 16 slice GE scanner Brivo 305, which was manufactured in 2015 and installed in 2016. In each centre, before the commencement of the experiment, information about the availability of previous QC test records being conducted on the instillation of the equipment and the subsequent routine QC tests being conducted on the scanners was known. CT water phantom for head consisting of three sections, each corresponding to a single scan plane (resolution block, contrast membrane and water bath sections) provided by the manufacturer of each machine at the time of the scanner installation for the purposes of QC tests, was used to conduct the procedure. CT number for water, field uniformity, noise, HCR and low-contrast resolution were tested using the same phantom. The manufacturer's guidelines provided in the manual for QC tests were strictly followed. With the aid of phantom holder, the phantom was placed on the table-top and aligned such that it was at the centre of the gantry in the axial, sagittal and coronal planes. This was achieved by aligning the axial positioning light to the circumferential line marking section 1, coronal light to the horizontal lines on either side of the phantom and sagittal light to the vertical line on the top of the phantom. Brain protocol was used in scanning the phantoms in all the centres included in the study.
On the monitor, about 2 cm equal-sized circular ROI was placed at the centre and at 12, 3, 6 and 9 O'clock positions, respectively, over uniformity sections of the phantom image. CT number for water and image noise was determined by taking the mean CT number at the central ROI and SD of each ROI, respectively. The mean pixel values for the four peripheral ROIs were noted as the field uniformity in each phantom. Six sets of bar patterns in HCR module were examined to determine the limiting resolution, by identifying the smallest bar pattern in which we are able to see and count all the five bars. Then, using the standard algorithm, the SD of the pixel values of the multiple bar pattern was determined which validated the subjective assessment of the HCR. The low-contrast section was used to choose the first barely visible hole with the sharp margin within a doped polystyrene membrane being identified and recorded. The difference in the mean CT number between the water and water plus plastic was also recorded.
| Results | | |
There was no availability of previous QC test records being conducted on the installation of the equipment and the subsequent routine QC tests being conducted on the scanners that participate in the study. The information was sourced from the radiographers operating the scanners.
[Table 1] shows the CT number for water, field uniformity, image noise, high and low contrast resolutions for CT scanners I, II, III and IV. Scanner I had CT number for water and image noise of 2.5 and 4.5, the high contrast resolution was at bar pattern 5, and the low contrast resolution was on hole-4. Scanner II had CT number for water and image noise of -0.34 and 8.74, the high contrast resolution was at bar pattern 6, and the low contrast resolution was at hole-4. Scanner III had CT number for water and image noise of -0.47 and 4.9, the high contrast resolution was on bar pattern 3, and the holes for low contrast resolution were not demonstrated. Scanner IV had CT number for water and image noise of -0.13 and 4.93, the high contrast resolution was at bar pattern 4, and the low contrast resolution was at hole-4. | Table 1: Computed tomography number for water, field uniformity, image noise, high- and low-contrast resolution for scanners I-IV
Click here to view |
[Figure 1] (a) shows the result for the test of the CT number for water, field uniformity and image noise (b) shows the result of the test for high contrast, the HCR bars were countable at 5 HCR block (c) shows the result of the tests for low contrast resolution, the margins were well demonstrated at hole 4. | Figure 1: Computed tomography images of quality control phantom in centre III. (a) Computed tomography number for water, field uniformity and image noise. (b) High-contrast resolution. (c) Low-contrast resolution
Click here to view |
[Figure 2] (a) shows the result for the test of the CT number for water, field uniformity and image noise (b) shows the result of the test for high contrast, the HCR bars were countable at 6 HCR block (c) shows the result of the tests for low contrast resolution, the margins were well demonstrated at hole 4. | Figure 2: Computed tomography images of quality control phantom in centre II. (a) Computed tomography number for water, field uniformity and image noise. (b) High-contrast resolution. (c) Low-contrast resolution
Click here to view |
[Figure 3] (a) shows the result for the test of the CT number for water, field uniformity and image noise (b) shows the result of the test for high contrast, the HCR bars were countable at 3 HCR block (c) shows the result of the tests for low contrast resolution, the block of the low contrast resolution was not demonstrated on the image (0). | Figure 3: Computed tomography images of quality control phantom in centre I. (a) Computed tomography number for water, field uniformity and image noise. (b) High-contrast resolution. (c) Low-contrast resolution
Click here to view |
[Figure 4] (a) shows the result for the test of the CT number for water, field uniformity and image noise (b) shows the result of the test for high contrast, the HCR bars were countable at 4 HCR block (c) shows the result of the tests for low contrast resolution, the margins were well demonstrated at hole-4. | Figure 4: Computed tomography images of quality control phantom in centre IV. (a) Computed tomography number for water, field uniformity and image noise. (b) High-contrast resolution. (c) Low-contrast resolution
Click here to view |
| Discussion | | |
The findings of the study show that CT scanner I, III and IV has passed CT number for water (0.3, −0.48, −1.36) and field uniformity tests because the HU at centre and peripheral ROI are within the acceptable limits ±3 and ±5 as shown in [Table 1]. This is an indication that the reconstruction algorithm that computes the CT numbers of the images is working correctly; therefore, good-quality diagnostic images with low radiation dose are expected from the scanner.[12] However, CT scanner II has failed the CT number for water test because the value obtained (−4.38) as shown in [Table 1] is above the acceptable limits of ±3. Possible cause of the CT number of water to be out of range is mis-calibration of the algorithm generating CT number.[5] The implication is that images produced by the scanner may lack accuracy which might lead to inaccurate diagnosis. These findings are contrary to the findings of the studies conducted by Sidi,[4] Oliveira et al.[8] and Khoramian et al.,[13] which showed that CT number for water and field uniformity tests to be within acceptable limits for all the scanners studied, even though the latter study by Sidi [4] was conducted in the same study area with the current study. The possible explanation to the disagreement with Sidi [4] is that the scanner losses its integrity possibly due to wear and tears. All the scanners failed noise test (4.5, 8.75, 4.90 and 4.93) because the measured values of SD of central ROI on uniformity phantom images were out of the acceptable limits (±3) as shown in [Table 1], and this can compromise the quality of the images produced by the scanners which might lead to wrong diagnosis. The possible causes of noise in CT images are matrix size (pixel size), algorithm, scattered radiation, number of detected photons, slice thickness, electronic noise (detector electronics) and object size.[5],[12],[14] The implication of this is that the noise reduces contrast resolution and might hide pathology with density similar to surrounding tissues. Most pathologists imaged in CT is seen in soft tissue such as liver, brain, kidneys and lungs, and this can lead to miss interpretation of the CT images.[12] This finding is contrary to that of Sidi [4] and Khoramian et al.[13] which showed only one CT scanner failed noise QC test in each of the study.
Only CT scanner III passed the HCR test as all the five bars were clearly seen at the block of 0.8 mm bar pattern (3rd bar pattern) as shown in [Figure 1]; SD of the pixel values of multiple bar pattern was determined which validated the subjective assessment of high contrast by the researchers. The CT scanner I, II and IV failed the HCR test because the HCR bars were countable at the 5, 6 and 4 HCR block as shown in [Table 1]. Respective ROI SD values were found to be 26.2, 31.01 and 28.31 as shown in [Figure 1], [Figure 2], [Figure 3], respectively, which are below the acceptable limit of 37 ± 4. Enlarged X-ray tube focal spot, excessive mechanical wear in gantry motion, misalignment of electromechanical components or detector failures could be the cause of HCR test failure.[5] The implication of failed HCR test is that the images produced by scanner I, II and IV may fail to clearly demonstrate the edges of structures, margins of tumors, small foreign bodies and small bony structures, and this will certainly interfere with diagnosis. This is contrary to the findings of Kodlulovich et al.[15] All the scanners involved in this study failed lowcontrast resolution test because the 4th hole was the smallest hole to be clearly identified within the lowcontrast insert of the CT scanners I, II and IV as shown in [Figure 2], [Figure 3], [Figure 4]. However, CT scanner III failed to demonstrate the low-contrast inserts entirely as shown in [Figure 1]. Possible causes of HCR failure are improper calibration, fault in the detector or the presence of impurities in the area of interest which may indirectly have an effect on CT QC parameters. The possible implication is that the poor resolution may not allow clear visualisation pathologic tissue changes, especially soft tissue low-density tumors, which are mostly found within the brain, kidney and liver.[12],[15] This finding is in agreement with the findings of Njiki et al.[16]
Phantom images obtained with CT scanner II demonstrate some evidence of ring artefacts as shown in [Figure 2] above (ring artefact is a CT phenomenon that occurs due to mis-calibration or failure of one or more detector elements in a CT scanner). This can sometimes simulate pathology. Usually, recalibrating the detector is sufficient to fix this artefact, although occasionally the detector itself needs to be replaced.[17],[18] It is important to note that CT scanner II failed the entire QC tests conducted in this study. Recalibration of algorithms may be able to solve problems associated with a CT number for water and field uniformity tests.[11]
| Conclusion | | |
Majority of the CT scanners in Kano metropolis passed the tests for the CT number for water and field uniformity but revealed a high failure of noise, HCR and low-contrast resolution tests. Urgent intervention of the servicing engineer on the CT scanners is strongly recommended for corrective maintenance of the equipment.
Recommendations
- Engineers should be invited to take immediate measures about the failed QC test
- Each facility should ensure full implementation of QC programme and all information relevant to QC should be appropriately kept for future use.
Financial support and sponsorship
Nil.
Conflicts of interest
There are no conflicts of interest.
| References | | |
| 1. | International Atomic Energy Agency. Quality Assurance Programme for Computed Tomography: Diagnostic and Therapy Applications No 19. Vienna: International Atomic Energy Agency; 2012.  |
| 2. | Lell MM, Wildberger JE, Alkadhi H, Damilakis J, Kachelriess M. Evolution in computed tomography: The battle for speed and dose. Invest Radiol 2015;50:629-44. |
| 3. | |
| 4. | Sidi M. Evaluation of quality control parameters in diagnostic computed tomography scanners in Kano metropolis, Nigeria. Indian J Appl Res 2016;6:205-7. |
| 5. | |
| 6. | |
| 7. | |
| 8. | |
| 9. | |
| 10. | General Electronics. Quality Assurance Phantom Manual: Technical Reference Manual. GE Healthcare; 2005. |
| 11. | Euclid S. Computed Tomography: Physical Principles, Clinical Applications, and Quality Control. St. Louis: Saunders Elsevier; 2009. p. 482. |
| 12. | Goldoost B, Ebrahimpoor M, Behrouzkia Z, Aghdam RZ, Refahi S. Assessment of water CT number, field uniformity and noise in diagnostics computed tomography scanners in Urmia metropolis, Iran. Int J Adv Biotech Res 2018;9:165-70. |
| 13. | Khoramian D, Sistani S, Firouzjah RA. Assessment and comparison of radiation dose and image quality in multi-detector CT scanners in non-contrast head and neck examinations. Pol J Radiol 2019;84:e61-7. |
| 14. | Guleng MD. Radiation Dose and Image Quality in CT-Evaluation of How Slice Thickness, Tube Current Modulation and Reconstruction Algorithms Affect Radiation Dose and Image Noise; 2018. Available from: https://ntnuopen.ntnu.no/ntnu-xmlui/handle/11250/2565890. [Last accessed on 2020 Mar 07]. |
| 15. | |
| 16. | Njiki CD, Manyol JE, Yigbedeck YE, Abou'ou DW, Yimele BC, Sabouang JF. Assessment of image quality parameters for computed tomography in the city of Yaoundé. OJORAD 2018;8:37-44. |
| 17. | Boas FE, Fleischmann D. CT artifacts: Causes and reduction techniques. Imaging Med 2012;4:229-40. |
| 18. | Adejoh T, Onwudiliora OE, Ukaji NF, Nzotta CC, Nwefuru OS, Aniagolu CA, et al. Analysis of artefacts from computed tomography of the head at Nnamdi Azikiwe University Teaching Hospital Nnewi. Niger J Med Imaging Radiat Ther 2015;4:6-10. |
[Figure 1], [Figure 2], [Figure 3], [Figure 4]
[Table 1]
|
Search Pubmed for