|Year : 2020 | Volume
| Issue : 1 | Page : 33-37
Assessment of image quality of plain abdominal radiographs at a northwestern Nigeria tertiary hospital
Hassan Maidugu Musa1, Suleiman Lawal1, Aliyu Galadima1, Abubakar G Mathew2, Aliyu Sadiya Musa1, Halimat Shadiat A. Umar1, Chibuike Onyesoh1
1 Department of Radiology, Ahmadu Bello University Teaching Hospital, Zaria, Nigeria
2 Department of Medical Radiography, University of Maiduguri, Maiduguri, Nigeria
|Date of Submission||17-May-2020|
|Date of Decision||30-May-2020|
|Date of Acceptance||22-Jun-2020|
|Date of Web Publication||25-Sep-2020|
Dr. Suleiman Lawal
Department of Radiology, Ahmadu Bello University Teaching Hospital, Zaria
Source of Support: None, Conflict of Interest: None
Background: Image quality has a significant effect on the accuracy of radiographic diagnosis, and without optimal image quality many vital information might be missed and that will affect patient management negatively. This work is aimed at assessing the image quality of plain abdominal radiographs, which is one of the most frequent emergency radiographic examinations done, using the radiographic technical parameters and the guidelines of image quality criteria set by the European Commission (EC) as the measuring criteria for the development of quality assurance program. Materials and Methods: A prospective survey of anteroposterior erect and supine plain abdominal radiographs of adult patients produced at Ahmadu Bello University Teaching Hospital from August 2019 to October 2019 was conducted assessing its quality based on the radiographic technical parameters and the guideline set by the EC for assessing image quality using a data capture sheet. Results: Plain abdominal radiographs of 92 adults comprising, 55.4% male, 44.6% female, 42.4% erect, and 57.6% supine, were evaluated. The finding of the study revealed that image quality criteria on radiographic technical parameters were optimal in 2%, suboptimal in 50%, and poor in 48%. The image quality based on the EC guideline result shows that only 3% of the reviewed plain abdominal radiograph were optimum, 63.0% were suboptimal, and 34% were poor. Furthermore, the study proves that gender has no significant effect (P = 0.985) on the image quality based on radiographic technical parameter, whereas image quality based on the EC guideline is significantly influenced by gender (P = 0.023). Conclusion: The quality of plain abdominal radiographs produced at this center during this period was abysmal; hence, rigorous quality control programs, familiarization of the radiographers with the guidelines for assessing image quality, and total replacement of obsolete equipment need to be instituted.
Keywords: Abdomen, image quality, radiographs
|How to cite this article:|
Musa HM, Lawal S, Galadima A, Mathew AG, Musa AS, A. Umar HS, Onyesoh C. Assessment of image quality of plain abdominal radiographs at a northwestern Nigeria tertiary hospital. J Radiat Med Trop 2020;1:33-7
|How to cite this URL:|
Musa HM, Lawal S, Galadima A, Mathew AG, Musa AS, A. Umar HS, Onyesoh C. Assessment of image quality of plain abdominal radiographs at a northwestern Nigeria tertiary hospital. J Radiat Med Trop [serial online] 2020 [cited 2023 Mar 26];1:33-7. Available from: http://www.jrmt.org/text.asp?2020/1/1/33/296114
| Introduction|| |
Plain abdominal radiograph is among the common radiographic examination that forms a part of medical assessment, particularly in accident and emergency departments. Plain abdominal radiographs play a vital role in assessing the level of patient preparation for contrast-based examinations involving the gastrointestinal tract and urogenital system; it is also used in demonstrating kidney stones and other forms of calcifications.,
Plain abdominal radiography can be readily accessed in low-income countries such as Nigeria within a short time, thereby assisting the referring physician in having a clue of the disease or even arrive at a definitive diagnosis.
Image quality is a broad term and its definition depends on different categories of people and the specific intended purpose. It can be studied physically (objectively) or by intuition (subjectively). For instance, a radiologist observing a radiograph may be concerned chiefly in its diagnostic worth; the radiographer, on the other hand, may only concentrate on how well the radiograph represents the various anatomical structures.,
Imaging process, characteristics of equipment, and imaging system variables selected by the operator determine the image quality of a radiograph. Thus, an ideal set of guidelines and parameters to describe image quality is expected to measure the effectiveness with which a radiograph can be used for its intended purpose.,, Studies of image quality are essential for optimization of the radiography processes in any clinical settings.,,
According to the American College of Radiology (ACR) guidelines on assessing the quality of abdominal radiography, all abdominal radiographs that are performed should include the pubic symphysis and both hemidiaphragms and extend sufficiently laterally to ensure that the entire bowel is included as well as both flanks.
In general, high-quality radiographs provide maximum diagnostic information by improving the chance for correct diagnosis and ultimately contribute to quality patient care., Guidelines have been set by the European Commission (EC) that defines the diagnostic requirements for a normal basic radiograph, specifying anatomical image criteria and important image details. The objectives of this study are to assess the quality of plain abdominal radiographs produced at Ahmadu Bello University Teaching Hospital (ABUTH), Zaria, based on radiographic technical parameters and the guidelines set by the EC on assessing image quality and to determine relationship between gender and image quality. This center as it stands now has no quality assurance program in place. Furthermore, no prior research in this area has been documented, hence necessitating the need for baseline data or reference point for feature quality assurance program or as guide for future studies. This study is therefore an initial program intended to verify the quality of abdominal radiographs in this facility and then its subsequent management.
Ho = Gender has no effect on the quality of plain abdominal image based on radiographic technical parameters and the guidelines set by the EC.
H1= Gender has an effect on the quality of plain abdominal image based on radiographic technical parameters and the guidelines set by the EC.
| Materials and Methods|| |
This prospective survey was conducted at ABUTH, northwestern Nigeria, involving plain abdominal radiographs of 92 adults, assessing the image quality based on radiographic technical parameters and the guidelines set by the EC for assessing image quality. The radiographs were obtained using General Electric Medical System X-ray machine, Model: MS-18S Manufactured: October 2004, Japan and Fujifilm Computed Radiography Model: CR-IR392 Manufactured: October 2015, Tokyo-Japan.
Ethical clearance was obtained from the health research ethics committee of the institution.
Data capture sheet includes patients sex, radiographic position, radiographic technical parameters (i.e presence of full patient identification, correct placement of anatomical marker, evidence of adequate collimation including both hemidiaphragms, pubic symphysis, and flanks laterally, evidence of no rotation to one side and artifacts), EC guidelines for assessment of image quality (i.e visualization of kidney, liver and psoas muscles outlines, area of coverage, image centralization and visually sharp reproduction of bones).
A radiograph is considered to be optimal if all the 5 radiographic technical parameters are met, suboptimal if 3–4 of the radiographic technical parameters are met, and poor if <3 of the radiographic technical parameters are met. Moreover, a radiograph is considered to be optimal if all the 6 parameters of EC guideline for image quality criteria are met, suboptimal if 3–5 of the criteria are met, and poor if only <3 of the criteria are met. AP supine radiograph can be performed as a standalone projection or as part of an acute abdominal series, depending on the clinical question posed, local protocol, and the availability of other imaging modalities. Pearson's Chi-square test was used to test hypothesis and P < 0.05 was considered statistically significant.
Technique of plain abdominal radiography
The gowned patient lies on the X-ray table with both shoulders and hips equidistant from the table for the supine radiograph, while in erect radiograph, he stands with ventral abdomen toward the image detector. Projection is anteroposterior at full inspiration for supine and posteroanterior for the erect. The centering point is the midsagittal (equidistant from each anterior superior iliac spine) at the level of the iliac crest and 5 cm above the level of iliac crest for supine and erect, respectively. Collimation includes the lateral abdominal walls, the diaphragm and lung bases, and inferior to the inferior pubic rami. Radiographic orientation is portrait, detector size of 35 cm x 43 cm was used, and exposure factors were 70–80 kVp and 30–120 mAs.
| Results|| |
Plain abdominal radiographs of 92 adult subjects were assessed, out of which 55.4% (n = 51) were male and 44.6% (n = 41) were female. Furthermore, 42.4% (n = 39) of these radiographs were erect, while 57.6% (n = 53) were supine.
As shown in [Table 1] and [Figure 1], image performance analysis based on radiographic technical parameters shows that only 2% of the images were optimal, while 50% were suboptimal and 48% were poor. The detail revealed patient identification to be adequate in only 4.3% (n = 4) and not adequate in 87.0% (n = 80) of the radiographs, and collimation shows 40.2% (n = 37) adequate and poor in 33.7% (n = 31). In 92.4% (n = 85) of the radiographs, anatomical maker was present, whereas it was absent in 7.6% (n = 7). Symmetrical presentation of structures (i.e., absent of rotation) was observed in 68.6% (n = 63) of the radiographs, while 31.4% (n = 29) showed presence of rotation. Artifacts were absent in 84.8% (n = 78) of the radiographs, whereas 15.2% (n = 14) were denuded by the presence of artifact.
|Table 1: Results of assessed radiographic technical parameters and European Commission guideline criteria|
Click here to view
[Figure 2] illustrates the image performance of the radiographs based on EC guideline for image quality criteria revealing only 3% of the radiographs had optimal image quality, the majority were either suboptimal or poor (63% and 34% respectively). [Table 1] also displays further in-depth analysis of the components, comprising 70.7% (n = 65) adequate area coverage, 12.6% (n = 18) not adequate, and 9.7% (n = 9) poor. The kidney outlines were present in 10.9%, the psoas muscles outline in only 7.6% the liver was outlined in 64.1% of the radiographs. The study also revealed that 83.7% (n = 77) of the radiographs had sharp presentation of the bones, whereas 16.3% (n = 15) were not sharp; 62.9% (n = 58) of the radiographs were centralize, whereas 37.1% (n = 34) were not centralized.
|Figure 2: Image performance based on European Commission guideline for image quality criteria|
Click here to view
[Table 2] shows the relative effect of gender on image quality based on radiographic technical parameters (A) and EC guideline (B). The former (A), revealed that only 1.9% of the radiographs of the male participants were optimal, 51% suboptimal, and 47.1% poor; a similar trend was observed in the female with 2.4% optimal, 48% suboptimal, and 48% poor radiographs, respectively. While the latter parameter(B), show 3.9% of the male participants' radiographs to be optimal, 74.5% sub optimal and 21.6% poor, also the female had 2.4% optimal, 48% sub optimal and 48% poor radiographs respectively.
|Table 2: Patients gender performance in relation to and radiographic technical parameter score (A) and European Commission guideline for image quality criteria (B), respectively|
Click here to view
[Table 3] shows the relationship between genders in the two image quality assessment guidelines. The radiographic technical criteria (A) showed no significant statistical difference between the two genders (Chi-square test, P = 0.985). This depicts that gender does not affect image quality based on radiographic technical parameters, therefore affirming the null hypothesis.
|Table 3: The correlation between gender and image quality base on radiographic technical parameter score (A) and European Commission guideline (B), respectively|
Click here to view
Whereas based on the EC guideline(B) significant statistical difference exist between the two gender (Chi-square test P = 0.023). This depicts that gender affects image quality based on these latter criteria, therefore rejecting the null hypothesis and hence favoring the alternate hypothesis.
| Discussion|| |
The study revealed that only 2% of the reviewed plain abdominal radiographs were optimal based on the radiographic technical parameters, i.e., the plain abdominal radiographs that met all the five radiographic technical parameters, 50% were suboptimal, and 48% were poor. This result though abysmal, its similar to the findings of Cheema at East Surrey Hospital, who reported inadequacy of plain abdominal X-ray produced with only 18% adequacy, but in contrast to Shettima et al who reported 50.6% adequacy. The variation of this study from the others may be attributed majorly to the grossly inadequate patient identification as observed, having 87.0% 'not adequate'. This is majorly because of absence of patient age in the request form, making most of the radiographs falling short of the five sets of criteria, which is a wide contrast to the findings of Egbe et al. who recorded only 6.7% 'not adequate' and 81.3% adequate patient identification.
Collimation, as one of the technical parameters observed during the study, showed 40.2% adequate, 26.1% not adequate, and 33.7% poor. Shettima et al. also found that beam collimation is the most common factor that affects the quality of the radiographs with only 18.8% having adequate collimation, 6.9% not adequate, and 74.2% had poor collimation due to misalignment of the beam or improper collimation during procedures under collimation of the x-ray beam resulting in irradiation of areas of non-interest and poor beam centring resulting in cut-off of area of interest are the cause of the “poor collimation” observed in this study.
Another technical parameter assessed was the presence of permanent anatomical marker, which shows 92.4% presence. This is in line with the findings of Shettima et al. and Egbe et al. who reported 91.5% presence of anatomical marker and 86% presence and correctly positioned anatomical marker on the radiographs, respectively.
Egbe et al. reported that 40% of the radiographs had rotation due to poor patient positioning and Shettima et al. reported that 52.3% of the radiographs had rotation. This study shows a similar finding but less proportion of radiographs, that is, 31.4% with rotation due to poor positioning. These narrow variations may be attributed to difference in patients' clinical condition as this contributes to their ability to maintain position.
An incidence of 40.8% radiographic artifacts was reported by Shettima et al. and the artifacts were mainly due to improper fixation resulting in thiosulfide deposit on the radiographs produced. This study reported a very low incidence of artifact, i.e., 15.2% and were mainly due to elastic bands, button's and hooks from patents' short-knicker. This is contrary to the earlier findings of Shettima et al. in both values and nature of artifacts. These differences could be accounted for by the use of computed radiography in this index study as against darkroom radiography in the former.
On the basis of EC guideline, this study revealed that 3% of the reviewed radiograph were optimal, 63% suboptimal, and 34% poor. This is in tandem with the study of Shettima et al. whose findings also shows 3.8% adequate, but contradict that of Egbe et al. who reported quite high adequate value of 80% in teaching hospitals, 62% in specialist hospitals, and 74.1% in private hospitals. This variation is attributed to the different scoring methods used; Egbe et al. considered the score of 3 and above to be adequate, while this index study only considers the score of 6 to be adequate.
In this study, the visualization of kidney and psoas muscles outline as components of the EC guideline revealed; 10.9% and 7.6% respectively. These low values can be attributed majorly to the patient's clinical condition at the time of examination, as most of the cases were acute abdomen with patient in painful distress and agitated, hence staying motionless for imaging is a difficult task. The resultant motion blurring contributes to the poor visualization of the kidneys and psoas muscles outline. This finding is in keeping with that of Shettima et al. who reported visualization of kidneys outline in 9.6% and psoas muscle outline in 16.9%.
However visualization of liver outline was in 64.1% of the radiographs which is in contrast to Shettima et al. who reported a paltry 19.2%. The use of film-screen radiography mainly by the latter as against computed radiography of the index study could account for this difference.
Sharp presentation of bones has the highest percentage of 83.7% as observed in this study, which is in agreement with the finding of Shettima et al. who reported visualization of bone in 78.1% and also as the parameter of image quality that has the highest percentage.
Furthermore, the study revealed that 62.9% of the reviewed radiographs were well centralized. This assertion is in line with the finding of Egbe et al. and that of Shettima et al. who reported 69.3% adequate patient positioning and 70.8% image centralization, respectively.
Shettima et al. reported that 64.2% of the reviewed radiographs were adequate in coverage, 21.2% were fairly covered, and 14.6% were poorly covered; Cheema reported that only 18% of the reviewed radiographs included the four anatomical land marks as described by the ACR. Findings of this study revealed 70.7% adequate coverage of the area of interest, 19.6% not adequate, and 9.8% poor, which is a wide disagreement with the result of Cheemah, but similar to that of Shettima et al.
In this study, gender did not significantly affect image quality when criteria for the assessment are radiographic technical parameter; however, a significant gender effect on the image quality was observed when EC guideline for image quality was considered. These variations might be attributed to the fact that all the radiographic technical parameters have little or nothing to do with the body physique of the individual than when compared to individual criteria for the EC guidelines. This study, first of its kind in this region, has brought to limelight the grossly inadequate\substandard radiographs produced at this facility and has revealed both the clinicians (not providing adequate clinical\biodata on request forms) and the radiographers (faulty radiographic techniques) to be culpable. The paucity of literature on this subject matter and lack of adequate information on subjects chronological age has made assessment of image quality in terms of age difficult and hence limited the discussion.
| Conclusion|| |
This study has unveiled that images produced at this facility are of suboptimal quality, which has the potential of causing missed or misdiagnosis, which could be of deleterious effect to the patient and frustrating to the clinician. An urgent need for improvements is desirable, and this can be achieved using this study as a baseline, through rigorous implementation of quality control programs, familiarizing the radiographers with the guidelines for assessing image quality, and total replacement of obsolete equipment.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Wadman M, Syk I, Elmstå hl B, Ekberg O, Elmståhl S. Abdominal plain film findings in acute ischemic bowel disease differ with age. Acta Radiol 2006;47:238-43.
Egbe NO, Eduwem DU, Ikamaise VC. Investigation of image quality of plain abdominal radiographs in three Nigeria hospitals. Biomed Imaging Interv J 2007;3:39.
Smitt JE and Hall EJ. Diagnostic imaging pathways indications of plain abdominal X-ray in emergency department. Emerg Med J 2009;26:160-3.
Pina DR, Duarte SB, Ghilardi Netto T, Trad CS, Brochi MA, de Oliveira SC. Optimization of standard patient radiographic images for chest, skull and pelvis exams in conventional x-ray equipment. Phys Med Biol 2004;49:N215-26.
International Commission on Radiological Protection, Radiological and Safety in Medicine: Publication 73, Annals of the ICRP. Oxford, UK: Pergamon Press; 1996. p. 26.
Sprawls P. Physical Principles of Medical Imaging 2nd
ed. Montreat, NC, USA;Sprawls Educational Foundation; 2014. Available from: http//www.sprawls.org/resources
. [Last accessed on 2020 May 15].
Egbe NO, Inah GB, Azogor WE, Chiaghanam NO, Ikamaise VC. Good radiographic practice: The Nigerian experience and the CEC recommendations. Europ J Radiog 2010;11:147-50.
Claudia R, Junior N, Catarina S, Ifrah A, Seraphine de H, Jose AG. Optimization of exposure parameters for spinal curvature measurements in pediatric radiography radiation dose and image quality optimization in medical imaging. Erasmus Int Prog Optimax Lisbon 2014;1:37-44.
InahGB, Akintomide AO, Edim, UU, Nzotta C, Egbe NO. A study of pelvic radiography image quality in a Nigerian teaching hospital based on the Commission of European Communities (CEC) criteria. South Afr Radiographer 2013;51:15-8.
Shettima AB, Malgwi FA, Gunda NM, Nwobi IC, Gajerima FL, Abubakar A, et al
. Evaluation of image quality of plain abdominal radiographs at university of Maiduguri teaching hospital, North-Eastern Nigeria. IOSR J Nurs Health Sci 2017;6:86-9.
European Commission. European Guidelines on Quality criteria for Diagnostic Radiographic Images. Luxembourg; 1996.
Whitely AS, Jefferson G, Holmes K, Sloane C, Anderson C, Hoadley G. Clarks Positioning Radiography. 13th ed. London: CRC Press; 2016. p. 365-83.
Cheema H. The quality of abdominal radiographs at east surrey hospital, United Kingdom. J Adv Radiol Med Imaging Abdominal 2016;1:203.
[Figure 1], [Figure 2]
[Table 1], [Table 2], [Table 3]