Download Estimation of Entrance Surface Dose and Image Quality Assessment

University of Ghana http://ugspace.ug.edu.gh
UNIVERSITY OF GHANA
COLLEGE OF BASIC AND APPLIED SCIENCES
ESTIMATION OF ENTRANCE SURFACE DOSE AND IMAGE QUALITY
ASSESSMENT OF ADULT PATIENTS UNDERGOING COMPUTED
RADIOGRAPHY EXAMINATIONS
BY
DANIEL ACKOM
(10507037)
THIS THESIS IS SUBMITTED TO THE UNIVERSITY OF GHANA, LEGON
IN PARTIAL FULFILLMENT OF THE REQUIREMENT FOR THE
AWARD OF MPHIL MEDICAL PHYSICS DEGREE.
DEPARTMENT OF MEDICAL PHYSICS, SCHOOL OF NUCLEAR AND
ALLIED SCIENCES
JULY, 2016
University of Ghana http://ugspace.ug.edu.gh
DECLARATION
This thesis is the result of research work undertaken by Daniel Ackom in the
Department of Medical Physics, University of Ghana, under the supervision of Prof.
Cyril Schandorf, Dr Stephen Inkoom and Mr. Edem Kwabla Sosu.
I hereby declare that this thesis is the result of my own original research and that no
part of it has been presented for another degree in this University or elsewhere. Duly
other works and/or researches done by other researchers cited in this work have been
acknowledged under references.
……………………………..
……………………………..
DANIEL ACKOM
PROF. CYRIL SCHANDORF
(STUDENT)
(PRINCIPAL SUPERVISOR)
Date ………………………..
Date.………………………..
……………………………..
……………………………...
DR STEPHEN INKOOM
MR EDEM KWABLA SOSU
(CO-SUPERVISOR)
(CO-SUPERVISOR)
Date ………………………..
Date…………………………
ii
University of Ghana http://ugspace.ug.edu.gh
ABSTRACT
The entrance surface dose (ESD) of radiographic examinations of adult patients
undergoing computed radiography (CR) examinations have been estimated and image
quality using ImageJ software version 1.48. In all 619 adult patients were involved in
the study comprising 243 males and 376 females. The mean ESDs calculated were in
the range of 0.29±0.0041 mGy to 6.08±0.55 mGy for chest PA, lumbar spine
AP/LAT, cervical spine AP/LAT, skull AP/LAT and abdomen AP. All the results
were lower compared with published results and diagnostic reference levels set by the
IAEA and UK show except chest PA examination which the value estimated at UGH
was by a factor of 2.56, 2.20 and 1.65 higher than the mean ESDs values by Inkoom
et al, the IAEA and Public Health of UK respectively. An increment by factor of 3.07,
2.87 and 2.15 were found in the estimated mean ESDs for chest PA examination at
KBTH for the published results respectively. Image quality of the radiographic images
was assessed in terms of CNR and SNR. The CNR for all examinations were between
3.54±3.27 and 20.63±8.65 for all examinations. A maximum difference 17.09 in
CNR was found between KBTH and UGH. The results obtained for SNR for both
hospitals showed that 92.86% of all the images assessed were at least 1.22 higher than
the Rose Model of the threshold value of ≥5. The study therefore has shown that the
estimated mean ESD of the hospitals were within the recommended references values
and the images were of good quality. Hence, there is the potential to reduce the dose
to patients while keeping images of diagnostic quality.
iii
University of Ghana http://ugspace.ug.edu.gh
DEDICATION
To my uncles Mr George Affran-Annan and Mr John Annan whose hands have been
a guide on my shoulders throughout my education.
To my Mother, Martha Annan, no parent could have done anymore.
To my siblings Bernard and Gloria Ackom, thank you for always being there for me.
iv
University of Ghana http://ugspace.ug.edu.gh
ACKNOWLEDGEMENTS
I wish to express my heartfelt gratitude to the persons whose guidance, enthusiastic
encouragement, sympathetic comments during the planning and development of this
study. I am indebted to my supervisors, Prof Cyril Schandorf, Dr Stephen Inkoom and
Mr Edem Kwabla Sosu who have been a constant source of words of encouragement
and motivation, valuable and constructive suggestions and guidance in the preparation
of this thesis.
This work would not have been successful without the help of Mr Prince Gyekye and
Mr Emmanuel Akrobeto of National Regulatory Authority. To the Authorities and
radiography staff at radiology departments of University of Ghana and Korle-Bu
Teaching hospitals, thank you all for your selfless support during gathering of data for
this study.
My dearest thanks to the Annan and Ackom family especially Mr John Annan, Mr
George Affran-Annan and his family, my mother, Martha Annan and Madam Joyce
Hutton Mills for their continual support and encouragement. My colleagues of 2016
medical physics class, may God bless you all for your good company throughout our
studies and I will never forget all the chats and beautiful moments we had together.
Special thanks to Madam Ethel Jobson-Mitchual and her family for their inspiration
and support throughout this work.
Finally, I would like to thank my dearest Sheila Ofosua Gyekye for her subservient
help, patience, support, care, hard work and for being there for me.
v
University of Ghana http://ugspace.ug.edu.gh
TABLE OF CONTENTS
DECLARATION .........................................................................................................ii
ABSTRACT ............................................................................................................... iii
DEDICATION ............................................................................................................ iv
ACKNOWLEDGEMENTS ......................................................................................... v
TABLE OF CONTENTS ............................................................................................ vi
LIST OF TABLES ....................................................................................................... x
LIST OF FIGURES .................................................................................................... xi
LIST OF PLATES ................................................................................................... xiii
LIST OF ABBREVIATIONS ................................................................................... xiv
CHAPTER ONE .......................................................................................................... 1
INTRODUCTION ....................................................................................................... 1
1.1. Statement of Problem ........................................................................................ 2
1.3. Relevance and Justification of the Study .......................................................... 4
1.5. Structure of the Thesis ...................................................................................... 5
CHAPTER TWO ......................................................................................................... 6
LITERATURE REVIEW............................................................................................. 6
2.1. Advancement in Radiography Systems ............................................................ 6
2.2. Entrance Surface Dose ...................................................................................... 9
2.2.1. Methods of Estimating ESD ...................................................................... 9
2.2.1.1. ESD Estimation by the use of Semiconductor Dosimeters ................... 10
vi
University of Ghana http://ugspace.ug.edu.gh
2.2.1.2. Measurement of ESD with TLD ........................................................... 13
2.2.1.3. Measurement of ESD with Ionization Chamber ................................... 16
2.2.1.4. Measurement of ESD with Dose Area Product Meter .......................... 18
2.3. Radiation Dose and Image Quality ................................................................. 19
2.4. Optimization.................................................................................................... 20
2.5. Image Quality Assessment with ImageJ Software in Medical Diagnostics .... 21
CHAPTER THREE.................................................................................................... 27
MATERIALS AND METHOD ................................................................................. 27
3.1. Study Area ....................................................................................................... 27
3.2 CR Systems used .............................................................................................. 27
3.3. The Ionization Chamber .................................................................................. 29
3.4. Quality Control (QC) on the CR systems ....................................................... 29
3.4.1. Tube Voltage Accuracy............................................................................ 30
3.4.2. Exposure Timer Accuracy........................................................................ 31
3.4.3 kVp, Timer and Exposure Reproducibility ............................................... 31
3.4.4 X-ray Beam Quality Test (Half Value Layer Determination) .................. 32
3.4.5 Current-Time Linearity ............................................................................. 33
3.4.6. Collimation and Beam Alignment ........................................................... 33
3.4.7. Tube Output Measurement ....................................................................... 34
3.4. Patient Data Collection ............................................................................... 34
3.5. Exposure Factors and Techniques ................................................................... 35
vii
University of Ghana http://ugspace.ug.edu.gh
3.6. ESD Estimation ............................................................................................... 36
3.7. Image Quality Assessment .............................................................................. 37
CHAPTER FOUR ...................................................................................................... 38
RESULTS AND DISCUSSION ................................................................................ 38
4.1. Quality Control (QC) on CR Systems ............................................................. 38
4.2. X-Ray Output Measurement ........................................................................... 40
4.3. Summary of Examinations .............................................................................. 43
4.4. ESD Estimation ............................................................................................... 44
4.5. Comparison of ESDs with other Published Results ........................................ 48
4.6. Image Quality Assessment .............................................................................. 50
4.6.1. Contrast-to-Noise Ratio ........................................................................... 51
4.6.2. Signal-To-Noise Ratio ............................................................................. 54
4.7. Optimization of Patient Protection .................................................................. 57
CHAPTER FIVE ........................................................................................................ 63
CONCLUSION AND RECOMMENDATION ......................................................... 63
5.1. Conclusion ...................................................................................................... 63
5.2 Recommendations ............................................................................................ 64
REFERENCES........................................................................................................... 66
APPENDIX A ............................................................................................................ 72
APPENDIX B ............................................................................................................ 74
QC RESULTS AT KBTH.......................................................................................... 74
viii
University of Ghana http://ugspace.ug.edu.gh
APPENDIX C ............................................................................................................ 77
QC RESULTS AT UGH ............................................................................................ 77
APPENDIX D ............................................................................................................ 79
ix
University of Ghana http://ugspace.ug.edu.gh
LIST OF TABLES
Table 2. 1: Timetable of developments in digital radiography .................................... 8
Table 4.1: Summary of QC results on CR at KBTH ................................................. 39
Table 4.2: Summary of QC results on CR at UGH .................................................... 40
Table 4.3: Estimated mean ESD at KBTH................................................................. 47
Table 4. 4: Estimated mean ESD at UGH .................................................................. 47
Table 4.5: Mean CNR calculated from images obtained for KBTH and UGH ......... 51
Table 4.6: Calculated SNR for the KBTH and UGH ................................................. 55
Table 4. 7: Optimized level of ESD and CNR ........................................................... 62
Table A1: Summary of individual patient exposures at KBTH ................................. 72
Table A2: Summary of individual patient exposures at UGH ................................... 73
Table B.1: kVp accuracy results ................................................................................ 74
Table B. 2: kVp, Exposure and Timer reproducibility............................................... 74
Table B. 3: mAs Linearity.......................................................................................... 75
Table B. 4: Timer Accuracy ....................................................................................... 75
Table B. 5: Tube Output measurement ...................................................................... 76
Table C. 1: kVp accuracy and Reproducibility results .............................................. 77
Table C. 2: Timer Accuracy and Reproducibility ...................................................... 77
Table C. 3: Tube Output measurement ...................................................................... 78
Table D.1: Datasheet used to collect exposure parameters at the hospitals ............... 79
x
University of Ghana http://ugspace.ug.edu.gh
LIST OF FIGURES
Figure 2.1: Diode dosimeter circuit configuration (b) Complete diode dosimeter .... 11
Figure 2.2: (a) Structure of MOSFET (b) Complete MOSFET dosimeter .............. 11
Figure 2.3: Setup for entrance diode calibration (b) Setup for exit diode calibration in
total body irradiation (TBI) condition (Allahverdi et al., 2008). ............................... 13
Figure 2. 4: TLDs in different forms (Radiation Products Design, 2014). ................ 14
Figure 2. 5: ROIs drawn to measure CNR (b) ROIs for SNR measurement ............. 23
Figure 2. 6: Main window of the ImageJ plugin developed for performing the
physical analysis and image quality checks. .............................................................. 25
Figure 2. 7: Schematic diagram of ROIs chosen. Source: (Song et al., 2004) ........... 26
Figure 3.1: Schematic diagram for different QC tests ............................................... 30
Figure 3.2: Schematic diagram of setup for measurement. ........................................ 35
Figure 4.1: Output curve of CR system at KBTH ...................................................... 41
Figure 4. 2: Output curve of CR system at UGH ....................................................... 42
Figure 4.3: Chart showing the percentage of each examination at KBTH ................ 43
Figure 4. 4: Chart showing the percentage of each examination at UGH.................. 44
Figure 4. 5: Chart comparing the mean ESDs from the two hospitals ....................... 46
Figure 4.6: Comparison of ESD at KBTH with published results and DRLs from
international organizations. ........................................................................................ 50
Figure 4. 7: Comparative bar chart of mean CNR obtained from KBTH and UGH . 52
Figure 4.8: Comparative bar chart of SNR obtained from the two hospitals ............. 56
Figure 4. 9: CNR as a function of ESD (mGy) for Chest PA .................................... 57
Figure 4. 10: CNR as a function of ESD (mGy) for Lumbar spine AP ..................... 58
Figure 4. 11: CNR as a function of ESD (mGy) for Lumbar spine LAT .................. 58
Figure 4. 12: CNR as a function of ESD (mGy) for cervical spine AP ..................... 59
xi
University of Ghana http://ugspace.ug.edu.gh
Figure 4. 13: CNR as a function of ESD (mGy) for cervical spine LAT .................. 59
Figure 4. 14: CNR as a function of ESD (mGy) for Skull AP ................................... 60
Figure 4.15: CNR as a function of ESD (mGy) for Skull LAT ................................. 60
Figure 4. 16: CNR as a function of ESD (mGy) for Abdomen AP ........................... 61
xii
University of Ghana http://ugspace.ug.edu.gh
LIST OF PLATES
Plate 3.1: CR system at the University of Ghana Hospital. ....................................... 28
Plate 3.2: Fujifilm IP cassette type CC ...................................................................... 28
Plate 3. 4: (a) Beam alignment tool (b) Collimator tool ............................................ 34
xiii
University of Ghana http://ugspace.ug.edu.gh
LIST OF ABBREVIATIONS
ALARA
As Low As Reasonably Achievable
AP
Anterior-Posterior
BSF
Backscatter Factor
CCD
Charge-Couple Detectors/Devices
CNR
Contrast-to-Noise Ratio
CR
Computed Radiography
DAP
Dose Area Product
DDR
Direct Digital Radiography
DQE
Detector Quantum Efficiency
DRL
Diagnostic Reference Level
ESAK
Entrance Surface Air Kerma
ESD
Entrance Surface Dose
FFD
Focus-to-Film Distance
FSD
Focus-to-Surface Distance
HIS
Hospital Information System
HVL
Half Value Layer
IAEA
International Atomic Energy Agency
ICRP
International Commission on Radiation Protection
IPEM
Institute of Physics and Engineers in Medicine
KAP
Kerma Area Product
KBTH
Korle-Bu Teaching Hospital
kVp
Kilovoltage Peak
LAT
Lateral
mAs
Milliampere seconds
xiv
University of Ghana http://ugspace.ug.edu.gh
MOSFET
Metal Oxide Semiconductor Field Effect Transistor
MTF
Modulation Transfer Function
NPS
Noise Power Spectrum
PA
Posterior-Anterior
PACS
Picture Archiving Communication Systems
QA
Quality Assurance
QC
Quality Control
RMI
Radiation Measuring Instrument
ROI
Region of Interest
SNR
Signal-to-Noise Ratio
SSD
Source-to-Surface Distance
TBI
Total Body Irradiation
TLD
Thermoluminescence Dosimeter
UGH
University of Ghana Hospital
WHO
World Health Organization
xv
University of Ghana http://ugspace.ug.edu.gh
CHAPTER ONE
INTRODUCTION
1.0. Background
X-ray was discovered in 1895 by a German experimental physicist, Wilhelm Roentgen
while working on emissions from electric current vacuum. He observed a glow from
barium platinocyanide (BaPt(CN)4) coated screen kept across the room whenever the
current was passed between the two electrodes in a charged cathode ray tube (Bansal,
2006). The first medical applications of x-rays was reported few months after the
discovery where a two dimensional projection image of patients anatomy was created
on a photographic plate using the specific attenuation properties of x-rays within
tissues (Jin Mo Goo et al., 2000). Since the discovery of x-rays, diagnostic imaging
has evolved and advanced to an extent that it has become an indispensable component
of patient diagnosis, management and in certain cases, treatment.
The type of x-ray systems used in medicine are conventional (computed or direct
digital radiography), dental, fluoroscopy, orthovoltage, mammography, megavoltage
and computed tomography. Conventional film-intensifying screen radiography has
been used for decades and has served the medical profession well. With the advent of
computers, digital imaging has now become the standard of human imaging (Mattoon
and Smith, 2004). Direct digital and CR by the use of flat panel detectors and reusable
phosphor screens, offer convenient and reliable way to replace the film. In CR, a
photostimulable phosphor plate is used for detection of x-rays instead of the
conventional film screen. The exposed plate is then scanned with helium neon laser
and the emitted light is captured by photomultiplier tube and converted to analogue
electrical system, which is then digitised. For direct digital radiography (DDR), a
1
University of Ghana http://ugspace.ug.edu.gh
semiconductor based sensor directly converts x-ray energy into electrical signals,
hence eliminating the middle step of latent image and image plate reader. Each of these
type of radiography system can either be manual or with automatic exposure control
system. Computed and direct digital radiography as at 2011, formed only 4% of the
conventional x-ray machines in Ghana to help improve the delivery of quality
healthcare system (Inkoom et al., 2011). They have the advantage of being low cost,
non-invasive, familiarity with medical professionals, relative harmlessness and fast
imaging times. However, there are reports showing that low doses of ionizing radiation
exposure encountered in diagnostic examinations may induce malicious conditions
(Kothan and Tungjai, 2011).
Use of x-ray facilities and equipment has increased rapidly in medical practices.
Diagnostic radiology has an enormous share of public dose from man-made sources.
With the advent of new digital imaging systems in radiography departments, it is
therefore important to focus on keeping the dose as low as reasonably practicable
whilst producing an image of diagnostic quality for digital radiography. The lack of
consistent feedback to technologists concerning the use of optimal acquisition
techniques is also a major problem with the use of computed and direct digital
radiography. This problem, along with the much larger dynamic range of digital
systems, has led to a gradual increase in patient radiation dose (Williams et al., 2007).
1.1. Statement of Problem
Image quality is one of the most important aspects of diagnostic radiology. The
concept of image quality has been undergoing a transformation with the widespread
use of digital-projection radiography. An ideal image is an image with high contrast,
high spatial resolution and low noise level (Vladimirov, 2010). However, these factors
2
University of Ghana http://ugspace.ug.edu.gh
are interconnected and depend on each other for every image. Since the introduction
of CR in Ghana in the year 2012, no attempt has been made to conduct the assessment
of the ESD to patient and associated image quality to trigger the need for optimization
of protection of patients. The current study seek to bridge this knowledge gap and
propose procedures for management of patient dose in CR and DR in Ghana. To fulfil
the Medical Exposure Directive (MED, 97/43, Euratom) requirement of a good image
quality at a radiation dose that is as low as reasonably achievable, optimization of
medical imaging systems is therefore deemed necessary (Smans et al., 2010). CR
systems are much more tolerant to inappropriate techniques with a good or even
perfect image. This is due to the high latitude of digital detectors and phosphor plates
which make possible systematic over-exposure or unnecessary high doses. Also, for
CR systems with automatic exposure control (AEC), until now, there is no certain
approach with regards to how to keep constant image quality at different tube
potentials in order to calibrate the AEC (Vladimirov, 2010).
1.2. Objectives
To address the above problem, this work seeks to:
 Accurately estimate the ESD of the patients (adults) undergoing selected CR
examinations [chest (PA), abdomen (AP), lumbar spine (AP and LAT), cervical
spine (AP and LAT) and Skull (AP and LAT)]
 Compare ESDs to the diagnostic reference levels established by international
organizations.
 Determine the potential for optimization of protection of patients for the selected
examinations under the study.
3
University of Ghana http://ugspace.ug.edu.gh
 Make appropriate recommendations from the findings
1.3. Relevance and Justification of the Study
This study would be helpful in the optimization of patient protection and setting
standards of good practice in the selected hospitals. All radiographic examinations are
performed purposely to provide reliable diagnostic information allowing rapid and
suitable treatment of the patient. These examinations must be performed with greater
care to ensure sufficient image quality while exposing the patient to the lowest dose
possible. According to literature, radiation dose to patients in diagnostic x-ray
examinations can be best estimated in terms of ESD per radiograph (Suliman and
Habbani (2007). To this point, the ESD is defined as the absorbed dose in air, including
the contribution from backscatter, assessed at a point on the entrance surface of a
specified object (IAEA: TRS-457, 2007). Literature also suggests that ESD is most
appropriate for simple x-ray projections since it meets the three (3) basic conditions
set out by the International Atomic Energy Agency (IAEA) which state that ESD is
simple to measure; permits direct measurement on patient during examination; and is
a representative of dose received by patients (Hamza et al., 2014).
1.4. Scope and limitation
This study was carried on the CR systems at the University of Ghana and Korle-Bu
Teaching hospitals in the Greater Accra Region. It focussed on some selected CR
examinations including chest (PA), lumbar spine (AP and LAT), cervical spine (AP
and LAT), skull (AP and LAT) and abdomen (AP).
4
University of Ghana http://ugspace.ug.edu.gh
1.5. Structure of the Thesis
The thesis is organized as follows: Chapter one presents the general background to the
thesis. It also presents the problem statement with set objectives to solve the problem
identified. The relevance, justification and scope of the study are also dealt with in
chapter one. Chapter two presents literature review of previous work done in relation
to this work. The step-by-step procedure followed to collect data for the study is
presented in chapter three. Chapter four deals with analysis of the data obtained. It
further presents the discussion of the various results obtained. Finally, Chapter five
summarizes the findings of the study, draws conclusions and gives recommendations
for further study and applications of the findings of this work.
5
University of Ghana http://ugspace.ug.edu.gh
CHAPTER TWO
LITERATURE REVIEW
This chapter reviews various literature in relation to this work in terms of ESD and
image quality of adult patients undergoing CR examinations. The chapter discusses
the various ways of determining the ESD and objective image quality assessment
using ImageJ.
2.1. Advancement in Radiography Systems
The discovery of x-rays by Wilhelm Roentgen a century ago was the first step to
enable the visualization of the human internal anatomical structures. This led to a
revolution in the field of medicine in particular with the use of x-rays for diagnosis
and treatment of different kinds of diseases and thus helping to improve healthcare
delivery in the world (IAEA: TRS-457, 2007). In the last 50 years diagnostic imaging
has grown from a basic application to a more advance level with the introduction of
different modalities. It is very clear that medical imaging has become established as
having an important role in patient management, and especially radiologic diagnosis
(Doi, 2006). Radiological facilities are found in even the smallest hospital and
emergency units involved in health care. A hospital without radiography is
unbelievable (Bansal, 2006). In the 1950s, most diagnostic images were acquired by
the use of screen–film systems and a high-voltage x-ray generator for conventional
projection x-ray imaging (Doi, 2006). Most radiographs were obtained by manual
processing of films in darkrooms, but with the passage of time, the latter is being taken
over by the use of automated film processing systems (Doi, 2006). Hence, during the
6
University of Ghana http://ugspace.ug.edu.gh
past two decades, CR has replaced the screen-film radiography in many radiological
facilities (Körner et al., 2007). Over the years, many significant modifications have
been made in the techniques and the equipment of radiography.
In 1977 a group of researchers first described the experimental digital subtraction and
was introduced into clinical use as the first digital imaging system in 1980 (Ovitt et al,
1980). X-ray images for general radiography were first recorded digitally with
cassette-based storage-phosphor plates and were also introduced in 1980 (Moore,
1980). In 1990, the first digital radiography (DR) system evolved and this was a
charge-couple devices (CCD) slot-scan system and subsequently in 1994
investigations of the selenium drum digital radiograph system were published (Neitzel
et al, 1994). The first flat-panel detector digital radiography systems based on
amorphous silicon as reported by Antonuk et al, (1995) and amorphous selenium
reported by Zhao and Rowlands, (1995) were introduced in 1995. The work of
Kandarakis et al, (1997) suggests that gadolinium-oxide sulphide scintillators were
introduced in 1997 and have been used for portable flat-panel detectors since 2001
(Puig, 2003). The latest development in digital radiography is the introduction of the
dynamic flat-panel detectors for digital fluoroscopy and angiography (Colbeth et al,
2001). Table 2.1 summarises the various developments made in digital radiography
systems over the years.
7
University of Ghana http://ugspace.ug.edu.gh
Table 2. 1: Timetable of developments in digital radiography (Körner et al., 2007)
Year
Development
1980
CR, storage phosphors
1987
Amorphous selenium-based image plates
1990
Charge-coupled device (CCD) slot-scan direct radiography (DR)
1994
Selenium drum digital radiography (DR)
1995
Amorphous silicon-cesium iodide (scintillator) flat-panel detector
1995
Selenium-based flat-panel detector
1997
Digital subtraction angiography
1997
Gadolinium-based (scintillator) flat-panel detector
2001
Gadolinium-based (scintillator) portable flat-panel detector
Dynamic flat-panel detector fluoroscopy-digital subtraction
2001
Angiography
2006
Digital Tomosynthesis
2009
Wireless DR (flat-panel detector)
Today, manufacturers provide a variety of digital imaging solutions based on various
detector and readout technologies (Körner et al., 2007). Of the many importance
derived from the use of digital radiography systems, the implementation of a fully
digital picture archiving and communication system (PACS) in which images are
stored digitally and are retrieved anytime cannot be ignored (Körner et al., 2007). With
the advent of PACS, image distribution in hospitals can now be achieved electronically
by means of web-based technology without loss of image fidelity. Körner et al., (2007)
further suggest that other advantages of digital radiography include higher patient
throughput, increased dose efficiency, and the greater dynamic range of digital
8
University of Ghana http://ugspace.ug.edu.gh
detectors with possible reduction of radiation exposure to patients. The dynamic range
of an x-ray imaging system is defined as the ratio of the largest and smallest input xray intensities that can be imaged (Williams et al., 2007). For all systems, the smallest
useful intensity is determined by the intrinsic system noise. The signal must be large
enough to exceed this noise, combined with the x-ray quantum noise. As we gradually
move towards the future of fully digital radiography systems, it therefore behooves
radiologists and other radiography technologists to be abreast with the technical
principles, image quality criteria, radiation exposure and other issues related to the
various digital radiography systems that are being produced (Körner et al., 2007).
2.2. Entrance Surface Dose
ESD is a quantity which depends on the x-ray penetrating power and parameters used
during the x-ray examination. It is evaluated because the dose to patient is greater at
the skin surface compared to the internal organ, indicating that radiation effect on the
skin is more significant (Fahmi, Aisyahton, and Zaky, 2013). The International
Atomic Energy Agency defines ESD as absorbed dose in air, including the
contribution from backscatter which is assessed at a point on the entrance surface of a
specified object (e.g. patient or phantom) (IAEA Human Health Series No. 24, 2007).
Research have shown that to estimate the amount of radiation in dose a patient receives
during x-ray examinations, the most commonly used quantity is the ESD (Fahmi et
al., 2013).
2.2.1. Methods of Estimating ESD
Radiation doses resulting from diagnostic radiological examinations are small and are
usually below the level of thresholds needed to cause deterministic effects. Exceptions
9
University of Ghana http://ugspace.ug.edu.gh
are found for interventional procedures in radiology and cardiology that may involve
high doses to the patient’s skin (IAEA TRS-457, 2007). The greatest source of
artificial ionization radiation exposure to the public or population is mainly from
diagnostic radiology and therefore, doses delivered in diagnostic radiological
procedures should be accurately determined in order to maintain a reasonable balance
between image quality and patient dose. There are various methods of estimating the
ESD of patients undergoing radiological examinations. In diagnostic radiology, ESD
can be estimated mostly by the use of common dosimeters for such as the ionization
chambers, TLD and semiconductor detectors.
2.2.1.1. ESD Estimation by the use of Semiconductor Dosimeters
Dosimeters other than ionization chambers exist and these have been used for
diagnostic measurements (IAEA:TRS-457, 2007). Silicon and Germanium detectors
which are semiconductor detectors have been used mainly for energy spectrometry
and have replaced scintillators in this application where highest energy resolution is
required. Some characteristics of semiconductor detectors also make them suitable to
be used as dosimeters for measuring dose or dose rate as a substitute for an ionization
chamber (Attix, 1986). In semiconductors with gap widths of 1 eV to 2 eV the number
of electron-hole pairs created by radiation in the vicinity of a junction between two
different materials is detected directly by electrical means. Hence, semiconductor
detectors work like solid state ionisation chambers. Examples of the semiconductor
dosimeters are the diode dosimeter circuit and complete diode dosimeter shown in
Figure 2.1 (a) and (b) respectively and also the metal oxide semiconductor field effect
transistor (MOSFET) dosimeter as shown in Figure 2.2 with the structure in (a) and
the complete MOSFET dosimeter in (b).
10
University of Ghana http://ugspace.ug.edu.gh
Figure 2.1: Diode dosimeter circuit configuration (b) Complete diode dosimeter
(Radiation Products Design, 2014)
Figure 2.2: (a) Structure of MOSFET (b) Complete MOSFET dosimeter (Cygler and
Scalchi, 2009)
With semiconductor detector, Owoade, Sambo, & Tijani, (2015) assessed the entrance
surface air kerma (ESAK) in patients undergoing chest x-ray from conventional
diagnostic radiology in Ogun State, Nigeria. The semiconductor device was used to
measure the output of the x-ray system and by applying equation 2.1 below, the
entrance surface air kerma for each patient was calculated.
Tube Output   mAs  BSF   FDD 
ESAK 
2
 FSD 
11
2
2.1
University of Ghana http://ugspace.ug.edu.gh
where mAs is the current-time product, BSF is the backscatter factor, FDD is the
focus-to-detector distance and FSD is the focus-to-surface distance. The results
showed significant differences in the ESAK of the various x-ray systems used. The
results compared well to similar studies from other countries except in chest
examination where the ESAK was higher than those in literature (Owoade et al.,
2015). Herath et al., (2015) obtained x-ray tube output measurement in assessing ESD
for postero-anterior erect chest x-ray examination of adult patients in a selected
teaching hospital in Sri Lanka. According to the authors, the ESD was then calculated
using equation 2.1 with FFD of 180 cm. The results from the study were below the
diagnostic reference level set by the IAEA. A study of diode calibration for dose
determination in total body was done by Allahverdi et al., (2008). Allahverdi et al.,
(2008) used four p-type diodes in connection to a MULTIDOSE electrometer as a
dosimeter for the measurement. The diodes were calibrated with water phantom of
dimension 30  30  32 cm3 . The setup for the study is shown in Figure 2.3. The
calibration factor was determined by these authors using the following equation:
Fcal 
absorbed dose ( D ) (measured with chamber )
diode reading ( M ) (measured with diode)
2.2
where Fcal is the calibration factor.
Allahverdi et al., (2008) found by measurement that the effect of angle incident on
diode response was significant and should have been taken into account. Variation in
thickness correction factor was found to be 0.7%.
12
University of Ghana http://ugspace.ug.edu.gh
Figure 2.3: Setup for entrance diode calibration (b) Setup for exit diode calibration in
total body irradiation (TBI) condition (Allahverdi et al., 2008).
Semiconductors have the advantage of being small, give instantaneous response to
irradiation, produce large signals from modest amount of radiation and are not
influenced by external factors like atmospheric pressure (IAEA TRS-457, 2007).
However, dose measurement with semiconductor detectors has some drawbacks
which include silicon having low atomic number (Z) of 14. This makes semiconductor
detectors to have large energy dependence by comparison with an ionisation chamber
(Maryanski et al, 1993). The detectors exhibit radiation damage and therefore, require
recalibration as the radiation sensitivity decreases with increasing radiation. Also, due
to their construction they show directional dependence, which can have significant
effects in some cases (Fong et al, 2001).
2.2.1.2. Measurement of ESD with TLD
Thermoluminescence is thermally activated phosphorescence; the most remarkable
and the most widely known of a number of different ionizing radiation induced
thermally activated phenomena. TLDs are a crystalline material that upon exposure to
ionising radiation are able to absorb or retain the radiation energy and store it at the
13
University of Ghana http://ugspace.ug.edu.gh
crystal lattice, known as electron traps (Khan, 1994). The use of TLDs in patient
dosimetry has been carefully considered in the publication of the European
Commission (Zoetelief et al., 2000). According to these authors, the availability of
TLDs in various forms (e.g. powder, chips, rods, ribbons) and made of various
materials. The availability of TL dosimeters in a variety of physical forms as shown
in Figure 2.4 below makes them especially suitable for skin dose measurements.
Figure 2. 4: TLDs in different forms (Radiation Products Design, 2014).
Shahbazi-Gahrouei, (2006) who measured ESD for routine x-ray examinations in
Chaharmahal and Bakhriari hospitals placed three TLD chips on the skin of each
patient and the doses were averaged for each radiography and mean of ESD of all
patients were calculated. According to this author, there was no significant difference
among the values of ESD obtained by the three detectors. However the ESDs were
higher that the diagnostic reference levels set by international organizations. The work
of Freitas & Yoshimura, (2009) on diagnostic reference levels for most of frequent
radiological examinations carried out in Brazil employed TLDs. According to the
authors, a pair of packed TLDs were attached directly to the patient’s skin at a point
close to the centre of the incident x-ray beam. The ESDs were calculated for only the
examination which showed relevant clinical information on the radiograph. The
14
University of Ghana http://ugspace.ug.edu.gh
results of the study compared to previous published data in Brazil showed a significant
difference even though different methods were used. A study of radiation exposure to
critical organs in the panoramic dental examination using TLDs was conducted on
different panoramic systems by Taghi et al, (2012). In this study, the absorbed dose to
organs and sensitive tissues in the head and neck region during panoramic radiography
were determined by placing TLDs packed in sachets on the thyroid gland, left and
right parotid glands, occipital region and the eye lids. The results according to these
authors show that there were differences between patient doses examined by different
panoramic systems. Mortazavi et al., (2004) also measured ESD on the thyroid gland
in orthopantomography by employing TLDs. The ESD was measured by placing three
TLD chips each on the thyroid of forty (40) patients undergoing panoramic
radiography examination. The doses received by each patients were averaged and the
mean ESD of all patients were then calculated. However, the results obtained
according these authors were inconsistent with the only reported data in literature
where the same method was employed. (Herath et al., 2015). In an assessment of dose
in thyroid and salivary glands in dental radiology using TLD a group of researchers
put a set five TLDs in the thyroid gland of an Alderson-Rando phantom and irradiated
it with 10 exposures using a set of exposure parameters provided by the equipment for
all techniques. The TLD response ( RTLDs ), in air kerma for a
137
Cs source, was
obtained by using the following formula:
RTLDs   LavgTLDs  BGavg   FN (mGy)
2.3
Where LavgTLDs is the average TLD readings, BGavg is the background radiation and
FN is the batch’s angular coefficient (Horizonte et al., 2011).
The main advantages of TLDs are tissue equivalent composition, smaller in size,
flexible in shape and are associated with good spatial resolution. They give a good
15
University of Ghana http://ugspace.ug.edu.gh
measurement accuracy when correctly used and are invisible in the x-ray image. TLDs
have also proven to be useful as personal dosimeters for radiation workers (Toivonen
et al., 2003). However, TLDs have a number of undesirable features which should be
considered before they are used. Toivonen et al., (2003) suggest that to measure the
dose received at a whole part of an exposed body, large numbers of TLDs are required
due to their small in size. Dose assessment can only be made retrospectively with TLD.
Dose assessment with TLD has also been shown to be laborious, capital intensive and
potentially intrusive when large patients are involved in the study (Mcparland, 1998).
Also, the main problem with TLDs is that there is typically a 0.1 𝑚𝐺𝑦 minimum
absorbed dose to produce a reasonably accurate resultThis value is above most of the
paediatric entrance surface air kerma values and thus makes the use of TLD in
paediatric dosimetry difficult (IAEA Human Health Series No. 24, 2013).
2.2.1.3. Measurement of ESD with Ionization Chamber
The ionization chamber is a device used for the detection and measurement of the
ionizing radiation. It does not measure radiation dose directly, instead it measures the
energy transferred to a small volume of air (kerma or exposure) by ionizing radiation.
It is the most widely used type of dosimeter for precise measurements of ionizing
radiation (Attix, 1986). Ionization chamber is used to measure kerma area product
(KAP) of an x-ray beam which is then converted to ESD. This procedure is referred
to as the indirect measurement of ESD. Incidence air kerma can be obtained with
suitable ionization chamber in the absence of phantom (IAEA TRS-457, 2007).
Kothan and Tungjai, (2011) estimated x-radiation output using mathematical model.
The authors further compared the calculated x-ray output with the measured output
using ionization chamber. The radiation output was measured at a source-to-chamber
16
University of Ghana http://ugspace.ug.edu.gh
distance of 100 cm. The results according to the authors shows a small difference
between the calculated and measured radiation outputs. Dose area product (DAP)
could be measured directly through the use of ionization chamber at the surface of xray tube collimator which can be converted into ESD using the equation
ESDDAP with CF 
DAP
 BSF  CFDAP
Area( FSD)
2.4
where FSD is the focus-to-skin distance, BSF is the backscatter factor and CFDAP is
the correction factor for the dose area product meter (Tamboul et al, 2014). The
authors opined that the results obtained by this method do not exceed the reference
values published from previous studies. The work of Wanai et al., (2011) who
estimated the ESD level of common diagnostic x-ray examinations at Songklanagarind
hospital made use of the ionization chamber to measure the entrance surface air kerma
of the x-ray equipment. The ESD was then estimated by the multiplication of the
entrance surface air kerma by an appropriate backscatter factor for each examination.
Here, the authors also suggests that most of the results obtained were in good
agreement of the diagnostic reference levels set by the IAEA except the skull lateral
which showed higher ESD compared to the corresponding reference level by the
IAEA. A feasibility study on the reduction of the ESD to neonates by the use of a new
digital mobile x-ray was done by Utsunomiya et al., (2013). The authors measured the
scattered radiation around the incubator with an ionization chamber survey meter. The
doses measured around the incubator were below measurable limits according to these
authors. The advantage of ionization chambers used in diagnostic radiology (unsealed
type) according to the IAEA is that no corrections for possible changes in air density
are required (IAEA TRS-457, 2007).
17
University of Ghana http://ugspace.ug.edu.gh
2.2.1.4. Measurement of ESD with Dose Area Product Meter
According to Kothan and Tungjai, (2011), there are various mathematical models
which can also be used to estimate the ESD but to employ the mathematical model,
the output of the x ray machine needs to be measured. This is because, the
mathematical model is a function of the output of the x-ray machine. The output of the
x ray machine can also be determine from a mathematical model (Kothan and Tungjai,
2011). The output is directly proportional to the peak tube voltage, 𝑘𝑉𝑝 , the tube
current, 𝑚𝐴𝑠, exposure time 𝑠 and inversely proportional to the square of the distance
(𝑑2 ) from the source. This is given as
Output (mR )  K  kV pn  mA  s 
1
d2
2.5
where, 𝑛 and 𝑘 are constant values but 𝐾 is the slope of the graph between 𝑚𝑅/𝑚𝐴𝑠
and 𝑘𝑉𝑝 2. The output is also dependent on the type of x-ray machine; single phase,
three phase or three phase and high frequency (Kothan & Tungjai, 2011). Hence, the
current pulse factor which also affects the radiation output of each type of x ray
machine should also be included in the mathematical model. The results obtained by
these authors agreed well with the measured results from the ionization chamber.
Fahmi et al, (2013) determined the ESD in chest radiography of anthropomorphic
phantom using two different indirect methods; the dose area product (DAP) with
correction factor ESDDAP with CF and the manual calculation method (ESDc). The
authors compared their results to the Standard Diagnostic Reference Level (DRL)
introduced by The European Union Council Directive 97/43/EURATOM and the
adoption of detector Correction Factor (CF) during ESD measurement. The
researchers used DAP meter to measure the dose per area and manually converted
18
University of Ghana http://ugspace.ug.edu.gh
DAP to the ESD value with an appropriate backscatter factor using the following
equations:
ESDDAP with CF 
DAP
 BSF  CFDAP
Area( FSD)
2.6
ESDDAP with CF 
DAP
 BSF
Area( FSD)
2.7
 FSD 
Area( FSD)  Area( FFD)  

 FFD 
2
2.8
Equation 2.6, and 2.8 were developed by Meade et al, 2003 and equation 2.7 was
derived by Livingstone et al, 2006 respectively. Fahmi et al, (2013) used an equation
derived by Osibote et al (equation 2.5) to calculate the ESD
2
 kVp 
ESDC  Output (OP)  
  mAs  BSF
 80 
2.9
The results of the authors showed that the ESDDAP with CF was found to be slightly
different compared to the Standard but not to a significant.
2.3. Radiation Dose and Image Quality
The optimization of x-ray imaging parameters remains a continuous challenge in
radiology since digital imaging provides new possibilities because of the wide
dynamic range of digital detectors and the digital image processing possibilities.
Aichinger et al., (2012) suggests that the knowledge of the relationship that links
image quality and radiation dose is a prerequisite to any optimisation of medical
diagnostic radiology. This is due to the fact that the ALARA principle requires that
the dose received by the patient undergoing radiological examination should be kept
19
University of Ghana http://ugspace.ug.edu.gh
“as low as reasonably achievable”. Also, European directive on usage of medical
exposures 97/43/Euratom (MED) requires that medical exposures have to be justified
and carried out in an optimized manner (Commission of the European Communities
(CEC), 1997). It further emphasizes that European Union member states shall adopt
appropriate criteria for radiological equipment in order to indicate when an action is
necessary, as well as, if appropriate, decommissioning the medical equipment
(Commission of European Communities (CEC), 2010).
According to Aichinger et al., (2012), the quality of the various components of the
imaging chain (focal spot, imaging geometry, image receptor, video camera and
amplifier, image-processing software, image display) has an overall effect on the
image signal obtained at the viewing station. The authors further opine that these given
facts must be taken into account when considering optimisation of image quality and
exposure. This is because image quality and radiation exposure cannot be investigated
independently of one another. Imaging procedures, including post processing, have to
be optimised to reach the necessary quality with lowest dose.
2.4. Optimization
New optimisation strategies have to be defined for digital radiography. It is important
that optimisation includes post processing which depends on different detectors and
exposure/dose parameters. Optimisation, a strategy of reducing dose to the patient
whilst still producing an image of diagnostic quality, is imperative in radiography
(Tugwell et al., 2014) and is therefore recommended by both the International
Commission on Radiological Protection (ICRP Publication 73, 1996) and the
European Medical Exposure Directive (CEC, 1997). This principle is important for all
20
University of Ghana http://ugspace.ug.edu.gh
examinations that involve ionizing radiation, however it is especially important for
high dose examinations. Furthermore, the physical characteristics of the x-ray tube
assembly and the radiographic device that are used in various radiological
examinations must be adapted to the medical requirements in such a way that the
image receptor generally can produce an optimum x-ray image with respect to the
representation of the object details which are needed for diagnosis at the lowest
possible level of exposure (Aichinger et al., 2012). These medical requirements can
be derived from national or international recommendations and guidelines. Aichinger
et al, (2012) further suggests that it is practical to optimize image quality first with
respect to medical indication since the whole-image information can be obtained from
the radiation image which exists in front of the imaging chain. This is based on the
assumption that theoretically, an ideal image receptor is being used and then taken into
account the real image receptor being employed. Image quality in medical imaging
systems can be described and quantified by three characteristics; contrast, noise and
sharpness. Contrast is mostly associated with screen film techniques whilst the derived
quantity SNR is mostly associated with digital imaging technique as an important
image quality parameter. Pascoal et al., (2006) states that the routine assessment and
control of image quality, both technical and clinical, is a fundamental task associated
with good practice. According to these authors, in addition to subjective visual
methods, currently, there are also available automated methods that can be used to
assess technical image quality associated with diagnostic imaging systems.
2.5. Image Quality Assessment with ImageJ Software in Medical Diagnostics
Image quality assessment still remains a challenge in the field of image processing. It
is still not satisfyingly solved and new approaches are still appearing. According to
21
University of Ghana http://ugspace.ug.edu.gh
the IAEA (IAEA: TRS-457, 2007), optimization processes which involves balancing
radiation dose and image quality in radiology does not always lead to reduction in
radiation dose and it is therefore important to emphasize that image quality must
always be sufficient to meet clinical requirements. It is essential to maintain the
appropriate level of image quality required for clinical confidence. Image quality
assessment plays an essential role in the various image processing applications. It
seeks to quantify a visual quality or, anatomically, an amount of distortion or
degradation in a given picture. These distortions (artifacts) are inevitably part of any
digital image processing chain from acquisition, processing and transmission of
images (Kud, 2012). A great deal of effort has been made to research which seeks to
develop various image quality metrics that correlate very well with the perceived
quality measurement but only limited success has been achieved (Wang, Bovik, and
Lu, 2002). The work of Mohammadi, Ebrahimi-moghadam, & Shirani, (2014)
suggests that the importance of efficient and reliable image quality evaluation has
increased due to the increasing demand for image-based applications. This according
to these authors is due to the fact assessing image quality is of fundamental importance
for numerous image processing applications where the goal of image quality
assessment methods is to automatically evaluate the quality of images in agreement
with human quality judgements. Generally image quality can be evaluated mostly by
using two approaches; objective and subjective methods (Kud, 2012). Subjective
image quality assessment is by visual inspection of the 2-D images by human observer.
Objective image quality assessment uses mathematical models to predict the quality
of an image accurately and automatically. ImageJ is an image analysis programme
created by the National Institute of Health, USA, that performs image quality
assessment objectively (Ferreira & Rasband, 2012). The authors opined that the
22
University of Ghana http://ugspace.ug.edu.gh
programme can display, edit, analyse, process, save and print 8-bit, 16-bit and 32-bit
images. The programme calculates the area and pixel value statistics of user-defined
selections. Image processing functions such as contrast manipulation, sharpening,
smoothing, edge detection and median filtering can be done with ImageJ software.
ImageJ can also calculate signal-to-noise ratio (SNR) and contrast-to-noise ratio
(CNR) by choosing one or more specific regions of interest (ROI) in the image. The
program uses one ROI for calculating the SNR and two ROIs to measure the CNR.
Ahmed et al., (2014) made a comparison of sonogram affirmed iterative reconstruction
and filtered back projection on image quality and dose reduction in paediatric head
computed tomography: a phantom study. The authors employed ImageJ to analyse the
images by determining the CNR and SNR of the images acquired. Two regions of
interest and three regions of interest were drawn on bone and soft tissue as shown in
Figure 2.5 (a and b) below to calculate CNR and SNR respectively.
Figure 2. 5: ROIs drawn to measure CNR (b) ROIs for SNR measurement (Ahmed et
al., 2014)
The CNR and SNR was then calculated by using the following equation:
CNR 
Mean signal of bone ( ROI 1)  Mean signal of tissue ( ROI 2)
S tan dard deviatio of ROI 2
23
2.10
University of Ghana http://ugspace.ug.edu.gh
SNR 
Mean signal value within ROI
S tan dard deviation within ROI
2.11
The results according to the authors showed that CNR remained constant within
minimal change in standard deviation. It also shows that Sinogram-Affirmed Iterative
Reconstruction enhances SNR while it reduces noise with possible reduction of dose
of 68% (Ahmed et al., 2014). In the work of Reis et al., (2014) on optimisation of
paediatrics CR for full spine curvature measurements using phantom: a pilot study,
the authors determined the CNR by using the ImageJ software. They also used two
ROIs and by employing equation 2.10 above, the CNR was calculated. The results
obtained showed a highest CNR at 66 kVp (C. Reis et al., 2014). Donini et al., (2014),
developed a free software for performing physical analysis of systems for digital
radiography and mammography. The software as shown in Figure 2.6 was used as a
plugin for ImageJ software. According to Donini et al., (2014), the plugin is able to
assist users in calculating various physical parameters such as the response curve also
known as the signal transfer property, modulation transfer function (MTF), noise
power spectrum (NPS) and detective quantum efficiency (DQE). This plugin can
further assist users to compute some image quality checks like defective pixel analysis,
uniformity, dark analysis and lag.
24
University of Ghana http://ugspace.ug.edu.gh
Figure 2. 6: Main window of the ImageJ plugin developed for performing the physical
analysis and image quality checks.
Donini et al., (2014) validated the software by comparing MTF and NPS curves on a
common set of images with those obtained with other similar programmes. The results
show a very good agreement (Donini et al., 2014). Song et al., (2004) report on
automated region detection based on the contrast-to-nose ratio (CNR) in the nearinfrared tomography. The authors employed ImageJ software to determine the CNR
by drawing regions of interest (ROI) on the image acquired as shown in Figure 2.7.
The CNR was then calculated by employing the following formula:
CNR 
 ROI  background
2
2
 wROI  ROI

 wbackground  background
2.12
where  ROI is the mean of the signal values in the target; background is the mean value
over the variable background,  ROI and  background are the standard deviations of the
25
University of Ghana http://ugspace.ug.edu.gh
target and of the background areas, respectively. Also the noise weights wROI and
wbackground are given by
wROI 
AreaROI
 AreaROI  Areabackground 
and wROI 
2.13
AreaROI
 AreaROI  Areabackground 
2.14
Figure 2. 7: Schematic diagram of ROIs chosen. Source: (Song et al., 2004)
Song et al., (2004) used the CNR as a basis for determining the detectability of objects
within reconstructed images from diffuse near-infrared tomography. The results show
that there was a maximal value of CNR near the location of an object within the image
and also the size of the true region could be estimated from the CNR (Song et al.,
2004). The work of Sun et al., (2012) on optimization of chest radiographic imaging
parameters: a comparison of image quality and entrance skin dose for digital chest
radiography systems also employed ImageJ software to estimate the noise levels. This
was done by placing ROIs of the same size and shape at different areas on the
background of the images acquired from the different systems. Significant differences
of image noise were found in the digital chest radiography from the different systems.
26
University of Ghana http://ugspace.ug.edu.gh
CHAPTER THREE
MATERIALS AND METHOD
This chapter presents the procedures that were followed to undertake this research. It
involves the description of the study area, the CR systems used, the type of ionization
chamber used and the types of test or examination that were performed using the set
of equipment available.
3.1. Study Area
This research was conducted in two major hospitals in Ghana within Accra in the
Greater Accra Region. These hospitals were the University of Ghana Hospital and
Korle-Bu Teaching Hospital (Room 5). These hospitals were chosen because of the
availability of CR systems since most of the hospitals in Ghana are still using the
conventional screen-film x-ray systems which employs dark rooms to process their
films.
3.2 CR Systems used
The CR system available at the University of Ghana Hospital was manufactured by
Philips Medical Systems, Japan in the year 2002. The system was installed at the
hospital in 2004. The tube head of the CR at this hospital has a model number
989000085271 with maximum kV of 150 kV, 300 mAs, 2.5 mmAl filtration at 75 kV.
Image of this CR system is presented in Plate 3.1. This CR serves the whole student
population of the university community and the general public. On an average, this
CR serves forty (40) patients per day but during the university clinical examination
27
University of Ghana http://ugspace.ug.edu.gh
for first year students, the number increases as high as one hundred and ninety (190)
patients per day.
At the Korle-Bu Teaching Hospital Room 5, the CR system being employed there was
manufactured by Shimadzu Corporation, Japan in the year 2012. The tube model is
53224558, 1.5 mmAl filtration at 70 kV and maximum kV of 150 kV. Averagely, this
CR serves forty (40) patients per day.
Plate 3.1: CR system at the University of Ghana Hospital.
All these systems employ Fujifilm cassette reader and IP standard cassettes type CC
shown in plate. 3.2 of dimensions 35×43cm, 35×35cm, 24×30m and 18×24cm
Plate 3.2: Fujifilm IP cassette type CC
28
University of Ghana http://ugspace.ug.edu.gh
3.3. The Ionization Chamber
The solid state ionization chamber employed in this study for the quality control and
the output measurements of the two CR systems was manufactured by Piranha RTI
(Piranha 657) with serial number CB2-11020219, dose range of 4 µGy/h - 273Gy/h or
0.4 mR/h – 31 kR/h, active detector area of 10×10 mm and was last calibrated on 5th
April, 2011.
3.4. Quality Control (QC) on the CR systems
Performance of each CR system was assessed through quality control. According to
Outif, (2004), the purpose of QC on x-ray system is to detect any change in the
performance of x-ray system, which may lead to an unacceptable image quality and/or
high dose to patient and staff. The setup for the QC is presented in Figure 3.1. The
following parameters were assessed during the quality control process:

Tube voltage accuracy

Timer accuracy

Voltage reproducibility

Exposure reproducibility

Half value layer

Current-time linearity

Collimation
29
University of Ghana http://ugspace.ug.edu.gh
Figure 3.1: Schematic diagram for different QC tests
3.4.1. Tube Voltage Accuracy
The beam was collimated to the sensitive part (10×10 mm) of the ionization
chamber.
To assess the kVp accuracy, the test was performed with variable kVp and constant
mAs. The set kVp ( kVpset ) on the console and the measured kVp ( kVpmeas. ) were
recorded. The percentage differences between kVpset and its corresponding kVpmeas.
were calculated using the equation
% Diff (kVp ) 
kVpset  kVpmeas.
 100
kVpmeas.
The results for this test for different x-ray systems.
30
3.1
University of Ghana http://ugspace.ug.edu.gh
3.4.2. Exposure Timer Accuracy
An accurate exposure timer is critical for proper exposure during examinations to
deliver a reasonable patient dose. Any significant variation from the desired exposure
will compromise the balance between image quality and patient dose as well as staff
radiation dose. The test was done with the same setup in Figure 3.1. Several exposures
with constant kVp and mA and with variable exposure time that cover the range of
possible exposure time of the x-ray system were performed. The set exposure time,
t set and measured exposure time, tmeas. were recorded and the percentage difference was
calculated using the formula:
% Diff ( Exposure time, ms ) 
tset  tmeas .
100
tmeas.
3.2
3.4.3 kVp, Timer and Exposure Reproducibility
The kVp, timer and exposure reproducibility tests also known as the coefficient of
variation (CV) was performed with constant kVp and constant mAs with the same
setup presented in Figure 3.1. Five sets of exposures were taken and in each case the
measured kVp, exposure (mGy) and exposure time (ms) were recorded. The
coefficient of variation for each parameter (kVp, exposure and exposure time) was
calculated using the following equation
CV 


3.3
where  is the estimated standard deviation of the five sets of measured parameter
(kVp, exposure and exposure time) and  is their corresponding mean for each
parameter.
31
University of Ghana http://ugspace.ug.edu.gh
3.4.4 X-ray Beam Quality Test (Half Value Layer Determination)
The quality of an x-ray beam is the ability of the beam to penetrate an object and this
is called the penetrating power. Hard x-ray beam is an x-ray beam with high
penetrability and x-ray beam with low penetrability is referred to as soft x-ray beam.
The quality of x-ray beam is expressed in terms of the half value layer (HVL). The
HVL is the thickness of a material required to reduce the intensity of an x-ray beam to
half of its original value. The test was performed with the same setup shown in Figure
3.1. One exposure was taken and the HVL was automatically given by the solid state
ionization chamber (Piranha 657) used. However this can be done manually by using
radiation measuring instrument (RMI) or meter. The meter is placed at a 100 cm
source-to-surface distance (SSD). The x-ray beam is collimated to the sensitive part
of the meter. With a set kVp and mAs combination, an exposure is taken three times.
Another set of exposure is taken by introducing aluminium attenuator 1 mm thick into
the primary x-ray beam on the collimated area of the meter. A varying thickness of
aluminium are added until the exposure falls to below 50% of the initial un-attenuated
value of the beam intensity.
A graph of exposure/current-time ratio as a function of aluminium thickness is plotted.
From the graph, the thickness of Al required to reduce the intensity of the x-ray beam
to its initial value is the half value layer (HVL) of the radiation beam.
The HVL can also be determined mathematically by using the following equation:
HVL 
ln 2
3.4

where  is the linear attenuation coefficient of the absorbing material. The results of
this test for the various CR systems employed.
32
University of Ghana http://ugspace.ug.edu.gh
3.4.5 Current-Time Linearity
The same setup shown in Figure 3.1 was used to perform this test. Several exposures
with constant kVp and variable mAs were taken. The output for each CR system was
recorded. Each output measured was divided by the corresponding mAs. The mAs
linearity variance (LV) was calculated using the following equation:
  Output / mAs    Output / mAs  
max
min
LV  

2  Output / mAs avg


3.5
where  Output / mAs max ,  Output / mAs min and  Output / mAs avg are respectively
the maximum, minimum and average values of the tube output per mAs recorded.
3.4.6. Collimation and Beam Alignment
The collimator and beam alignment test tools shown in plates 3.4 (a and b) were used
to assess the alignment of the x-ray field and how perpendicular the x-ray beam to the
image receptor. The collimator tool was placed at 100 cm SSD on the x-ray table under
the x-ray tube. The beam alignment tool was placed at the centre of the collimator tool.
The collimator shutters were adjusted in such a way that the edges of the light field
coincide with the rectangular outline on the collimator tool. That is, when an image of
the test tool is generated, the image of the two steel balls overlaps only when the beam
is accurately perpendicular to the image receptor. If an x-ray field falls within the
image of the rectangular frame, there is a good alignment. If an edge of the x-ray field
falls on the first spot, 1.0 cm , on either side of the line, it shows that the edges of the
x-ray field are misaligned by a percentage of 1%.
33
University of Ghana http://ugspace.ug.edu.gh
Plate 3. 3: (a) Beam alignment tool (b) Collimator tool
3.4.7. Tube Output Measurement
This test was also performed using the setup presented in Figure 3.1. Several exposures
were taken with variable tube voltage (40 kVp to 150 kVp with interval of 10 kVp)
and variable mAs (50 mAs to 2 mAs). A ratio of exposure/current-time product was
calculated for each kVp set. A graph of exposure/current-time product (mGy/mAs) as
a function set tube voltage (kVp) was plotted. From this graph, the output (mGy/mAs)
of each kVp used for each type of examination was calculated .
3.4. Patient Data Collection
To collect various information needed to estimate the ESD of each type of
radiographic examination considered, a data sheet was designed. The data sheet
captures information such as area of examination (e.g. Chest), patient age, sex, patient
thickness, applied tube voltage (kVp), current-time product (mAs) and focus-to-film
distance (FFD) for each type of examination. The data sheet was used to record patient
data and technique factors for each examination done. Patient thickness was measured
using a pair of callipers. Focus-to-film distance was measured by employing tape
34
University of Ghana http://ugspace.ug.edu.gh
measure (5 m long). The setup for each examination is presented in the schematic
diagram in Figure 3.2
Figure 3.2: Schematic diagram of setup for measurement.
3.5. Exposure Factors and Techniques
Exposure parameters and patient data were collected for a total of 619 patients with
243 males representing 39.3% and 376 females representing 60.7% for the two
hospitals. The minimum and maximum age of the patients that formed part of this
study were 18 and 82 years. This age group is considered adult due to the fact that it
is within the adult age category by the International Commission on Radiological
Protection (ICRP publication 101a, 2006). Summary of results of ESDs estimated,
exposure parameters and patient information for individual patient exposures are
35
University of Ghana http://ugspace.ug.edu.gh
presented in Tables A1 and A2 in Appendix A. For a good radiographic technique for
the projections chosen , the European guidelines on quality criteria for diagnostic
radiographic images requires 125 kVp, 180 cm FFD for chest; 70 – 90 kVp, 115 cm
FFD for lumbar spine AP; 80 – 100 kVp, 115 cm FFD for lumbar spine LAT; 70 -85
kVp, 115 cm FFD for skull AP/PA and LAT; 70 kVp, 115 cm for cervical spine AP/PA
and LAT; and 70 -100 kVp, 115cm for abdomen (European Commission, 1996).
3.6. ESD Estimation
The ESD received by each patient undergoing each type of examination considered in
this study was calculated by using the following equation (IAEA: TRS-457, 2007) :
ESD  ESAK  BSF
3.6
where ESAK is the entrance surface air kerma given by
 100 
ESAK  Output (mGym A s )  mAs  
 FSD 
1
1 1
2
3.7
FSD is the focus-to-skin distance which is obtained by subtracting the patient
thickness from the focus-to-film distance. That is
FSD  FFD  t p
where
3.8
t p is the patient thickness. The output measurement of each patient
examination is obtained from the output/current-time ratio (mGy/mAs) versus tube
voltage (kVp) graph. A backscatter factor of 1.35 was used. This is because according
to the European commission, (1996), backscatter factors vary between 1.3 and 1.4 for
the x-ray qualities used for various projections included in the quality criteria except
for mammography. Therefore, a single average value of 1.35 can be used in most cases
without appreciable error. The results were then compared to the diagnostic reference
36
University of Ghana http://ugspace.ug.edu.gh
levels (DRLs) set by international organisations like the International Atomic Energy
Agency and Public Health of United Kingdom.
3.7. Image Quality Assessment
A sample of ten images were acquired for each type of examination from each CR
system employed. Image quality was assessed using ImageJ software (version 1.48).
This is an image processing software developed by National Institutes of Health. It
works by using objective physical characteristics of the imaging system such as spatial
resolution, contrast and noise. The images were assessed by measuring the CNR and
the modulation transfer function (MTF) which determines the spatial resolution of the
imaging system.
To calculate CNR, two ROIs were marked on the images using a bespoke software
(ImageJ). ROI1 was applied mid-way of the vertebral body (maximum density) and
ROI2 in the lung region with a homogenous density (minimum density). The formula
in equation 3.9 was used to the estimate CNR:
CNR 
Mean signal of bone ( ROI 1)  Mean signal of tissue ( ROI 2)
S tan dard deviatio of ROI 2
3.9
The results obtained were then compared to the kVp at which each examination was
performed and the corresponding ESD.
The SNR was determined by drawing two ROIs; one on area of higher density and the
other on the background of the image. The SNR was then calculated using equation
3.10:
SNR 
Mean signal value within ROI
S tan dard deviation within ROI
37
3.10
University of Ghana http://ugspace.ug.edu.gh
CHAPTER FOUR
RESULTS AND DISCUSSION
This chapter presents results of the study and discussions of the findings arising from
the analysis of the results. The results and discussions cover; quality control, x-ray
beam output, patient ESDs, image quality (CNR and SNR) measurements.
4.1. Quality Control (QC) on CR Systems
To obtain a good balance between the dose received by patients undergoing
radiographic examinations and the corresponding image quality, the assessment of
physical operations of the x-ray systems is very essential. Results obtained from
assessing the CR systems performance are presented in Tables 4.1 and 4.2 for KBTH
and UGH respectively. The applied voltage across x-ray tube during examination has
effect on beam quality, image quality and the corresponding patient dose. The
accuracy and reproducibility of the tube voltage selected on the control console is
therefore important for a proper exposure technique selection. Table 4.1 and 4.2 show
that the tube voltage accuracy and reproducibility of the CR systems were within the
acceptable limit (≤ ± 6.0 % and ≤ 0.05) set by the Institute of Physics and Engineers
in Medicine. The exposure timer accuracy was also within the acceptable limit of ≤
±10% set by the Institute of Physics and Engineers in Medicine (IPEM) (IPEM Report
91, 2005). This shows that the x-ray tube generators were capable of terminating
exposure after the preselected time interval has elapsed. The kVp, timer and exposure
reproducibility for both CR systems were also below the limit of ≤ 0.05 set by IPEM
(IPEM Report 91, 2005). Therefore, the results show that the performance of CR
systems were consistent.
38
University of Ghana http://ugspace.ug.edu.gh
Table 4.1: Summary of QC results on CR at KBTH
Acceptable
Parameter
Deviation of
Deviation by
CR system
(IPEM)
Remarks
Tube Voltage Accuracy
-1.900%
≤ ± 6.000 %
Pass
Timer Accuracy
4.490%
≤ ±10.000%
Pass
0.001
≤ 0.050
Pass
0.011
≤ 0.050
Pass
0.002
≤ 0.050
Pass
Tube Voltage Reproducibility at
80kVp (Coefficient of Variation)
Timer Reproducibility at 80kVp
(Coefficient of Variation)
Exposure Reproducibility at
80kVp (Coefficient of Variation)
Half Value Layer at 70 kVp
Current-time Linearity
3.230 mm Al
0.026
≥ 2.100 mm Al Pass
≤ 0.100
Pass
The half value layers of the CR systems as observed in Tables 4.1 and 4.2 were
respectively 3.23 mmAl and 5.05 mmAl at KBTH and UGHs. Comparing these values
to the acceptable limit of ≥ 2.10 mmAl shows that the x-ray beams were of good
quality. Hence, much of the x-rays with lower energies which do not form part of
image but contributes to patient dose were eliminated. The two hospitals considered
did not have a quality assurance (QA) and QC programmes in place. This shows that
with proper QA/QC programmes in place, there is the potential to achieve dose
optimisation and patient protection with good or even perfect image quality.
39
University of Ghana http://ugspace.ug.edu.gh
Table 4.2: Summary of QC results on CR at UGH
Acceptable
Deviation of
Parameter
deviation by
Remarks
CR system
(IPEM)
Tube Voltage Accuracy
-1.150%
≤ ± 6.000 %
Pass
Timer Accuracy
-5.250%
≤ ±10.000%
Pass
0.006%
≤ 0.050
Pass
0.014
≤ 0.050
Pass
0.032
≤ 0.050
Pass
Tube Voltage Reproducibility at
80kVp (Coefficient of Variation)
Timer Reproducibility at 80kVp
(Coefficient of Variation)
Exposure Reproducibility at
80kVp (Coefficient of Variation)
Half Value Layer at 81 kVp
5.050
Current-time Linearity
0.051
≥ 2.100 mm Al Pass
≤ 0.100
Pass
4.2. X-Ray Output Measurement
Figures 4.1 and 4.2 show the relationship between the x-ray output (mGy/mAs) and
the applied tube voltages (kVp) at a constant tube current-time product (mAs) of the
CR systems used in this study. R2 values of 0.9856 and 0.9971 observed from the
output curves of Figures 4.1 and 4.2 show that there exist a strong correlation between
the parameters measured and thus show a good fit. The nature of the graph is also in
good agreement with the graph obtained from output versus kilovoltage obtained from
the work of Tamboul et al., (2014).
40
University of Ghana http://ugspace.ug.edu.gh
0.12
Output (mGy/mAs) = 2E-06kVp2.2066
R² = 0.9856
0.1
Tube Output (mGy/mAs)
0.08
0.06
0.04
0.02
0
0
20
40
60
80
100
Tube Voltage (kVp)
Figure 4.1: Output curve of CR system at KBTH
41
120
140
160
University of Ghana http://ugspace.ug.edu.gh
0.06
Output (mGy/mAs) = 2E-05kVp1.7118
R² = 0.9971
0.05
Output (mGy/mAs)
0.04
0.03
0.02
0.01
0
0
20
40
60
Set voltage (kVp)
Figure 4. 2: Output curve of CR system at UGH
42
80
100
120
University of Ghana http://ugspace.ug.edu.gh
4.3. Summary of Examinations
Ionizing radiation is used daily in hospitals and clinics to perform diagnostic imaging
procedures. The largest source of exposure to the Ghanaian population is due to
artificial sources with more than 95% resulting from diagnostic x-ray examinations
(Inkoom et al., 2012). This is due to the common use of x-rays in conventional,
computed and direct digital radiography systems. Figure 4.3 and 4.4 present the
summary of the number of each examination. It was observed from Figures 4.3 and
4.4 that 34 % of all radiographic examinations conducted in these hospital were chest
examinations. This confirms the fact that medical chest x-ray examination is the most
common radiological examination performed in Ghana today.
Abdomen, 10%
Skull LAT, 6%
Chest PA, 34%
Skull AP, 6%
Cervial spine LAT,
9%
Cervial spine AP,
8%
Lumber spine LAT,
14%
Lumber spine AP,
13%
Figure 4.3: Chart showing the percentage of each examination at KBTH
Chest x-ray examination forms 46.4% of all x-ray examinations in Ghana (Schandorf
and Tettey, 1998) and 40 % of all x-ray examinations conducted worldwide (WHO,
43
University of Ghana http://ugspace.ug.edu.gh
2016). It is also evident from the Figures 4.2 and 4.3 that skull x-ray is the least
requested examination which forms on 6 % of all the examinations considered for each
of the hospitals. This is due to the fact that skull examination is primarily conducted
in these hospitals on accident patients to determine whether there is a fracture in the
skull and then also most head (skull) scans are done using computed tomography.
Abdomen, 7%
Skull LAT, 6%
Skull AP, 6%
Chest PA, 33%
Cervial spine LAT,
12%
Cervial spine AP,
12%
Lumber spine AP,
12%
Lumber spine LAT,
12%
Figure 4. 4: Chart showing the percentage of each examination at UGH
4.4. ESD Estimation
Figure 4.5 compares the distribution of the mean ESDs obtained from the hospitals.
Tables 4.3 and 4.4 in addition to mean ESD, present 1st quartile, 3rd quartile, maximum
and minimum ESDs and standard deviation values of means of the ESDs for each
anatomical projection considered with their standard deviations. The highest mean
44
University of Ghana http://ugspace.ug.edu.gh
ESD of 3.01 mGy and 6.08 mGy were estimated for lateral projection of lumbar spine
for KBTH and UGH respectively. There was a difference of 3.07 mGy for lumbar
LAT projection between the two hospitals which shows that the mean ESD recorded
at UGH was higher than that of KBTH by a factor of 2.02. Lumbar spine AP projection
recorded the second highest mean ESD for UGH which was by a factor of 3.57 higher
the corresponding mean ESD recorded for KBTH. The third highest mean ESD was
recorded at UGH for abdomen also by a factor of 1.70 was higher than that recorded
at KBTH. However, the mean ESDs recorded for skull AP projection for both hospitals
were almost the same with a difference of 0.04 mGy. A lower mean ESD of 0.29 mGy
was recorded for cervical spine AP and LAT projections at KBTH. The values of mean
ESD were lower by factors of 1.76 and 2.07 as compared to that recorded at UGH for
both AP and LAT projections respectively. For chest examination, there was a
difference of 0.1 mGy between the mean ESDs recorded at KBTH and UGH. This
shows that the mean ESD recorded at KBTH was by a factor of 1.30 higher compared
to that of UGH of 0.33±0.01 mGy. Comparing the 1st quartile values for chest
examinations in the two hospitals shows that the dose recorded at KBTH was by a
factor of 1.19 higher than the dose recorded at KBTH. The 3rd quartile values were as
the same as the mean values were the dose recorded at UGH was lower than the dose
recorded at KBTH by a factor of 1.30.
45
University of Ghana http://ugspace.ug.edu.gh
6.08
Mean ESD (mGy)
3.5
3.01
2.62
1.7
1.74
1.54
0.98
0.33
Chest PA
0.53
0.43
Lumber
spine AP
0.29
Lumber
spine LAT
0.72
0.6
0.51
Cervical
spine AP
Cervical
spine LAT
UGH
KBTH
0.29
Skull AP
Skull LAT Abdomen
Figure 4. 5: Chart comparing the mean ESDs from the two hospitals
46
University of Ghana http://ugspace.ug.edu.gh
Table 4.3: Estimated mean ESD at KBTH
Area of Examination
ESD (mGy)
1st quartile Mean 3rd quartile Max Min StdDev
Chest PA
0.38
0.43
0.43
0.71
0.12
0.1022
Lumbar spine AP
0.80
0.98
1.05
1.57
0.63
0.2594
Lumbar spine LAT
2.91
3.01
3.05
4.19
2.46
0.2447
Cervical spine AP
0.29
0.29
0.30
0.30
0.29
0.0042
Cervical spine LAT
0.29
0.29
0.30
0.30
0.29
0.0041
Skull AP
1.77
1.74
1.83
1.87
1.19
0.1945
Skull LAT
0.52
0.53
0.54
0.54
0.52
0.0090
Abdomen
1.32
1.54
1.39
3.49
1.27
0.6107
*Max is maximum, Min is minimum and StdDev is standard deviation
Table 4. 4: Estimated mean ESD at UGH
Area of Examination
ESD (mGy)
1st quartile Mean 3rd quartile Max Min StdDev
Chest PA
0.32
0.33
0.33
0.36
0.30
0.01
Lumbar spine AP
3.39
3.50
3.65
3.99
3.03
0.23
Lumbar spine LAT
5.81
6.08
6.32
7.12
4.50
0.55
Cervical spine AP
0.49
0.51
0.53
0.59
0.44
0.03
Cervical spine LAT
0.56
0.60
0.63
0.71
0.51
0.05
Skull AP
1.67
1.70
1.73
1.79
1.65
0.04
Skull LAT
0.72
0.72
0.73
0.74
0.71
0.01
Abdomen
2.50
2.62
2.70
2.96
2.45
0.15
47
University of Ghana http://ugspace.ug.edu.gh
Table 4.3 also shows that a patient received a maximum dose as 0.71 mGy during
chest examination at KBTH as compared to 0.36 mGy of UGH for the same type of
examination. In general, Figure 4.4 further shows that all the mean ESDs recorded at
UGH were higher than their corresponding mean ESDs recorded at KBTH except for
chest PA and skull AP projections. The differences in the doses may be attributed to
the different exposure factors and techniques being employed by the radiographers at
the hospitals. The difference in the doses may also be due to variations in the number
of examinations per day and the thickness of the patients being examined.
4.5. Comparison of ESDs with other Published Results
The mean ESDs obtained from this study have been compared to other published data
and diagnostic reference levels (DRLs) set by international organizations (Public
health of UK and IAEA).
Figure 4.6 shows a bar chart for graphical comparison of the two hospitals. The Figure
shows that the highest means ESD was recorded for the lateral projection of lumbar
spine examination for both hospitals but they were however lower than their
corresponding mean ESDs from literature. At UGH, the mean ESD was lower than
that of Inkoom et al., (2012) by a factor of 1.61, 1.64 and 2.47 for UK (Public Health
England, 2016) and IAEA (Rehani, 2001) respectively. For the same lumbar LAT
projection, KBTH recorded a lower value by a factor of 3.24, 3.32 and 4.98 for Inkoom
et al., UK and IAEA respectively. Figure 4.6 further shows that all the mean ESDs
for the various examinations were lower than their corresponding published data and
DRLs except for chest examinations in which the ESDs (0.43 mGy for KBTH and
0.33 mGy for UGH) obtained from this study for both hospitals were higher than their
corresponding published data and reference levels by UK and IAEA. There was a
48
University of Ghana http://ugspace.ug.edu.gh
difference of 0.19 mGy, 0.18 mGy and 0.13 mGy between the estimated mean ESD
for chest examination at UGH and Inkoom et al., (2012) and the DRLs set by UK
(Public Health England, 2016) and IAEA (Rehani, 2001) respectively. The difference
shows that the mean ESD recorded at UGH was 135.71%, 117.86 % and 63.40 %
higher than the ESDs obtained by Inkoom et al., and the DRLs set by the Public health
of UK and the IAEA respectively. A percentage increment of 207.14%, 188.92 % and
116.68 % were found in the estimated mean ESD for chest examination at KBTH for
Inkoom et al., (2012), UK and IAEA respectively. This differences was due to the
selection of exposure parameters and technique factors including the coning of the xray beam. Other reason for the large variation in the mean ESD for chest may be due
to patient size, suboptimal usage of the equipment or equipment malfunctioning
generally because of the scarceness of regular quality control and radiation protection
programme. These differences suggest that much can be done to reduce the ESDs by
adequate changes of physical parameters without loss of image quality. Therefore, the
radiology departments of the hospitals should undertake a review of their radiographic
practice in order to bring their doses to optimum levels.
49
University of Ghana http://ugspace.ug.edu.gh
18.00
16.00
14.00
Mean ESD (mGy)
12.00
10.00
8.00
6.00
4.00
2.00
0.00
Chest PA
UGH
Lumber
spine AP
Lumber
spine LAT
KBTH
Cervical
spine AP
Inkoom et al., 2012
Cervical
spine LAT
UK (2016)
Skull AP
Skull LAT
Abdomen
IAEA (2001)
Figure 4.6: Comparison of ESD at KBTH with published results and DRLs from
international organizations.
4.6. Image Quality Assessment
Image quality is one of the most important aspect of radiology. The aim of radiology
is to acquire images which contain adequate information for clinical purpose with
minimum radiation dose to the patient. For optimization to be achieved, image quality
assessment must be made to balance the patient dose. To assess the image quality of
50
University of Ghana http://ugspace.ug.edu.gh
radiographic examinations, CNR and SNR have been calculated using values
generated from ImageJ software version 1.48.
4.6.1. Contrast-to-Noise Ratio
CNR is a measure for assessing the ability of an imaging procedure to generate
clinically useful image contrast. CNR was calculated for ten (10) radiographs for each
type of anatomical projection considered. The mean and standard deviation values of
the calculated CNR have been presented in Table 4.5. Figure 4.7 presents the graphical
representation of the mean CNR calculated for both hospitals.
Table 4.5: Mean CNR calculated from images obtained for KBTH and UGH
Mean CNR
Area of Examination
KBTH
UGH
kVp
KBTH
UGH
Chest
12.8 + 5.77
6.15 +3.14
62.0 - 68.0
125
Cervical spine AP
13.64 + 4.11
10.89 + 3.47
60.0 - 62.0
66.0 - 70.0
Cervical spine LAT
19.18 + 6.78
12.89 + 4.56
60.0 - 62.0
66.0 - 70.0
Lumbar spine AP
11.48 + 8.01
8.97 + 3.6
64.0 - 74.0
77.0 - 85.0
Lumbar spine LAT
3.79 + 11.97
4.81 + 3.51
80.0 - 95.0
96.0 - 98.0
Skull AP
7.43 + 3.02
9.89 + 3.87
73.0 - 75.0
75.0 - 78.0
Skull LAT
20.63 + 8.65
3.54 + 3.27
62.0 - 64.0
70.0 - 74.0
51
University of Ghana http://ugspace.ug.edu.gh
25.00
20.00
CNR
15.00
10.00
5.00
0.00
Chest
Cervical
spine AP
Cervical
spine LAT
Lumber
spine AP
KBTH
Lumber
spine LAT
Skull AP
Skull LAT
UGH
Figure 4. 7: Comparative bar chart of mean CNR obtained from KBTH and UGH
The results in Table 4.5 and Figure 4.7 show that the highest CNR of 20.63 + 8.65 was
obtained for skull LAT projection at KBTH where the image was acquired at a kVp
range of 62 to 64 kVp as against 3.54 + 3.27 where the examination was done at a kVp
range of 70 to 74 kVp for the same projection at UGH. The CNR recorded for skull
LAT projection was also the lowest among all the projections considered. A difference
of 17.09 between CNR for skull LAT projection accounted for the CNR of KBTH
52
University of Ghana http://ugspace.ug.edu.gh
being higher than that recorded at UGH by a factor of 5.83. The second highest CNR
was recorded for cervical LAT projection at KBTH where the CNR recorded was by
a factor of 1.48 higher than that recorded at UGH. A similar higher CNR was recorded
at KBTH over their corresponding values obtained at UGH for chest PA, cervical AP,
and lumbar AP. In all the three examinations, the kVp used to acquire images at
KBTH were lower than their corresponding kVp at UGH as shown in Table 4.6.
However, UGH recorded a higher CNR of 13.32 + 9.37 as compared with 4.63 + 0.48
recorded at KBTH for skull AP projection even though the range of kVp used for this
examination was higher at UGH than kVp used at KBTH for the same examination.
The CNR recorded for skull AP examination at KBTH was by a factor of 2.88 lower
than the CNR recorded at UGH. The same trend of higher CNR was recorded for
lumbar spine lateral projection at UGH as compared to CNR recorded at KBTH. Here
also, the kVp range used to acquire the image at UGH were higher than that used at
KBTH according to table 4.6. Generally, Figure 4.7 and table 4.6 show that the CNR
obtained at KBTH for all the examinations were higher than the CNR recorded at UGH
except for skull AP and lumbar spine LAT projection where higher CNR was recorded
at UGH as against KBTH. The variation in CNR between the two hospitals is mainly
due to the difference in kVp used for the various examination. A similar result was
obtained by Hess and Neitzel, (2012) who determined the CNR for paediatric
extremities using kVp range of 40 kVp to 60 kVp. According to the authors, the
highest CNR was obtained at 40 kVp. Also, the use of higher kVp is a well-known
strategy to reduce dose on paediatrics, but at the same time decreases the CNR (Reis
et al., 2014). The results of this study supports the assertion that the use higher kVp
generally decreases the CNR. This is because with lower kVp, there is less scatter
radiation reaching the detector with increasing mAs. Contrast is the most significant
53
University of Ghana http://ugspace.ug.edu.gh
factor influencing the choice of tube potential for imaging different parts of the human
body with different attenuations. The CNR obtained for objects containing several
hundred pixels like radiograph has the potential to provide a useful parameter for
comparing imaging performance for x-ray beams with different beam qualities. The
relationship between the CNR measurements and tube potential in general reflects the
variation in the number of details that can be detected (Doyle, 2008).
4.6.2. Signal-To-Noise Ratio
One of the most common indicator of image quality is the SNR. The ability to detect
an object (and hence resolve it from its neighbour) is related to the signal to noise ratio
(SNR) of the object. The common way to quantify the level of noise in an image is to
estimate the SNR. SNR has been calculated for ten images which were used to
determine the CNR. Table 4.6 and Figure 4.8 present the results obtained. The results
presented in Table 4.6 and Figure 4.8 show that the highest SNR of 22.35± 0.86 was
recorded for chest PA examination at KBTH as against 9.10 ± 0.81 SNR of the same
anatomical projection at UGH. This shows that SNR recorded at KBTH was 146%
more than the SNR recorded at UGH. The second SNR was also recorded at KBTH
which was by a factor of 1.07 higher than the corresponding SNR recorded at UGH
for cervical spine AP anatomical projection. A similar trend of KBTH having higher
SNR than the SNR recorded at UGH was found for lumbar spine AP/LAT projections
where the SNRs recorded at UGH were by factors of 1.43 and 1.11 lower than the
SNR recorded at KBTH for lumbar AP/LAT projections respectively.
54
University of Ghana http://ugspace.ug.edu.gh
Table 4.6: Calculated SNR for the KBTH and UGH
Mean SNR
Area of Examination
KBTH
UGH
Chest
22.35 ± 0.86
9.10 ± 0.81
Cervical spine AP
17.99 ± 7.67
16.81 ± 7.1
Cervical spine LAT
12.92 ± 3.93
17.58 ± 10.75
Lumbar spine AP
15.48 ± 1.27
10.85 ± 8.32
Lumbar spine LAT
13.59 ± 8.00
12.28 ± 3.46
Skull AP
4.63 ± 0.48
13.32 ± 9.37
Skull LAT
6.10 ± 3.60
14.17 ± 9.99
However, higher SNRs were recorded at UGH over KBTH for cervical LAT and skull
AP/LAT projections which accounted for 36.07%, 187.69% and 132.30% increase in
SNR for cervical LAT, skull AP/LAT respectively at UGH over the SNR at KBTH.
To be able to detect objects in a medical image, the threshold SNR is  5 (Bath, 2010).
This threshold was developed by Rose Albert who was interested to find the threshold
value of SNR of an object to be visible by human observer and this is known as the
Rose Model. The assumption of the Rose Model is that the factor that limits the
detection of an object in a radiographic image is the radiation dose used to produce
the image (IAEA: STI/PUB/1564, 2014). Comparing the results obtained to the Rose
Model show that 92.86 % of the images assessed had SNR greater than 5 for both
hospitals except the SNR recorded for skull AP projection at KBTH which had a value
lower than 5. This shows that clinicians would be able to extract useful information
from these images and there is the potential of reducing the dose patient received
undergoing the examinations considered.
55
University of Ghana http://ugspace.ug.edu.gh
25.00
20.00
SNR
15.00
10.00
5.00
0.00
Chest
Cervical
spine AP
Cervical
spine LAT
Lumber
spine AP
KBTH
Lumber
spine LAT
Skull AP
Skull LAT
UGH
Figure 4.8: Comparative bar chart of SNR obtained from the two hospitals
The differences in the SNR is due to the kVp and mAs factors used for the
examination. This is because the selection of exposure parameters such as the increase
of kVp and mAs result in more signal reaching the detector that should reduce the
noise in the image and improve the SNR.
56
University of Ghana http://ugspace.ug.edu.gh
4.7. Optimization of Patient Protection
In order to ensure dose and image quality optimisation for the various examinations,
the CNR was plotted as a function of the ESD for all examinations as shown in Figures
4.9 to 4.16. Optimisation technique employed in a radiology departments usually
consists of selections of tube potential, filtration and method of scatter removal and
there is a need to find an image quality parameter which can be used in clinical imaging
tasks to compare and evaluate different options. CNR has been examined as such a
parameter in this study. For a radiograph to be acceptable for diagnosis, it is dependent
on its ability to display the correct anatomical part being imaged with optimum levels
of CNR.
14
12
10
CNR
8
6
4
2
0
0
0.1
0.2
0.3
0.4
ESD (mGy)
Figure 4. 9: CNR as a function of ESD (mGy) for Chest PA
57
0.5
0.6
University of Ghana http://ugspace.ug.edu.gh
18
16
14
12
CNR
10
8
6
4
2
0
0.6
0.8
1
1.2
1.4
1.6
1.8
ESD (mGy)
Figure 4. 10: CNR as a function of ESD (mGy) for Lumbar spine AP
7
6
5
CNR
4
3
2
1
0
2.5
2.7
2.9
3.1
3.3
3.5
3.7
3.9
4.1
ESD (mGy)
Figure 4. 11: CNR as a function of ESD (mGy) for Lumbar spine LAT
58
4.3
University of Ghana http://ugspace.ug.edu.gh
20
18
16
CNR
14
12
10
8
6
4
0.284
0.286
0.288
0.29
0.292
0.294
0.296
0.298
0.3
ESD (mGy)
Figure 4. 12: CNR as a function of ESD (mGy) for cervical spine AP
20
19
18
CNR
17
16
15
14
13
12
0.2
0.22
0.24
0.26
0.28
0.3
ESD (mGy)
Figure 4. 13: CNR as a function of ESD (mGy) for cervical spine LAT
59
0.32
University of Ghana http://ugspace.ug.edu.gh
16
15
14
13
CNR
12
11
10
9
8
7
6
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
ESD (mGy)
Figure 4. 14: CNR as a function of ESD (mGy) for Skull AP
9
8.5
8
7.5
CNR
7
6.5
6
5.5
5
4.5
4
0.35
0.36
0.37
0.38
0.39
ESD (mGy)
Figure 4.15: CNR as a function of ESD (mGy) for Skull LAT
60
0.4
0.41
University of Ghana http://ugspace.ug.edu.gh
12
11
10
CNR
9
8
7
6
5
4
1
1.5
2
2.5
3
3.5
4
ESD (mGy)
Figure 4. 16: CNR as a function of ESD (mGy) for Abdomen AP
In general, the relationship between the CNR measurements and tube potential reflects
the variation in the detectability of objects in the image. From figures 4.9 to 4.16, the
level of CNR as a measure of image quality as well as low dose to patient has been
summarized in the Table 4.7. When the values of CNR and ESD presented in Table
4.7 are implemented in the clinical setting, the ESD for chest PA examination can be
reduced by 26.9 % at UGH and 65.4%
61
University of Ghana http://ugspace.ug.edu.gh
Table 4. 7: Optimized level of ESD and CNR
Type of Examination
CNR
ESD (mGy)
Chest PA
11.60
0.26
Lumbar spine AP
6.40
0.84
Lumbar spine LAT
3.60
3.26
Cervical spine AP
17.60
0.30
Cervical spine LAT
19.00
0.24
Skull AP
15.60
0.92
Skull LAT
4.60
0.38
Abdomen AP
10.65
0.26
62
University of Ghana http://ugspace.ug.edu.gh
CHAPTER FIVE
CONCLUSION AND RECOMMENDATION
This chapter provides the conclusions drawn from this study and recommendations
on the estimation of ESD and image quality assessment of adult patients undergoing
CR examination in Ghana.
5.1. Conclusion
The mean ESD estimated for all the examinations considered under this study was in
the range of 0.290±0.004 mGy to 6.08±0.55 mGy which was due to the different
exposure parameters and technique factors used for various examinations in the two
hospitals considered. The exposure parameters used for the examinations considered
were generally higher at UGH as compared to the exposure parameters used at KBTH.
Mean ESD of 6.08±0.55 mGy was obtained for lumbar LAT anatomical projection at
UGH was the highest among all the ESDs estimated. However, a lower mean ESD of
0.290±0.004 mGy was recorded for cervical AP/LAT projection at KBTH.
The mean ESD obtained were at least lower compared to other published results and
DRL set by international organization except the mean ESD for chest examination
which was higher than the other published results.
The CNR calculated was in the range of 3.53±3.27 to 20.63±8.65. The results showed
that the CNR obtained for KBTH were generally higher than the CNR obtained at
UGH. The differences were mainly due to the use of different kVp for the
examinations. It was also found that lower kVp increases CNR which in turn decreases
63
University of Ghana http://ugspace.ug.edu.gh
patient dose. The lowest CNR was recorded for skull LAT projection at UGH whilst
the highest CNR was also recorded for the same projection at KBTH.
The SNR was calculated as a means of quantifying the level of noise in the images
acquired from the CR systems used. There was a higher SNR of 22.35 ± 0.86 for chest
PA examination at KBTH and a lower SNR of 4.63 ± 0.48 which also occurred at
KBTH for skull AP projection. The results obtained was compared to the Rose model
where the limit of SNR was set at  5 . All the results obtained were greater than 5
except skull AP examination at KBTH. The results therefore showed that the CR
systems produce images that contain useful clinical information. The values of the
CNR and SNR can be used as the baseline for future quality control monitoring and
research. The plot of CNR against the ESDs for all the examinations showed that there
is a potential to reduce doses to patients while keeping images of diagnostic quality.
5.2 Recommendations
Hospital Authorities
There should be a comprehensive QA/QC programme in place and this should include
training of staff in effective use of equipment and approved imaging protocols.
Continuous checks and maintenance culture of the CR systems must be established.
There should be standardized imaging protocols established to be used uniformly by
radiographic staff in order to avoid the risk of unnecessary exposure to patients and
staff.
The findings of this work should be used to develop institutional level optimization
of protection of patients consistent with clinically acceptable image quality.
64
University of Ghana http://ugspace.ug.edu.gh
Further Work
Further research should be conducted to extend work to optimization of patient dose
while keeping images of diagnostic quality acceptable to radiologist and referring
clinicians. Future studies should be extended to other hospital that uses CR systems
for examination and this should include patient dose to paediatric patients who are
more sensitive to radiations than adults.
Regulatory Authority.
The regulatory authority should provide applicable regulations and guidance
document on radiation protection and safe use of CR in Ghana. In collaboration with
the relevant professional bodies assist in establishing diagnostic reference levels for
all examinations performed using CR systems in Ghana.
65
University of Ghana http://ugspace.ug.edu.gh
REFERENCES
Ahmed, A., Garcia, A., Bakker, A., Tomkinson, D., Salamin, J., and Lange, R. De.
(2014). A comparison of Sinogram Affirmed Iterative Reconstruction and filtered
back projection on image quality and dose reduction in paediatric head CT : a
phantom study. OPTIMAX 2014 - Radiation Dose and Image Quality
Optimisation in Medical Imaging, 27–36.
Aichinger, H., Dierker, J., Barfuß-Joite, S., and Sabel, M. (2012). Radiation Exposure
and Image Quality in X-Ray Diagnostic Radiology, Physical Principles and
Clinical Applications (Second). New York: Springer.
Allahverdi, M., Geraily, G., Esfehani, M., Sharafi, A., and Shirazi, A. (2008). Diode
calibration for dose determination in total body irradiation. Iranian Journal of
Radiation Resources, 6(1), 43–49.
Antonuk L. E., Yorkston, J., & Huang, W. (1995). A real-time, flat-panel, amorphous
silicon, digital x-ray imager. RadioGraphics, 15, 993-1000.
Attix, F. H. (1986). Introduction to Radiological Physics and Radiation Dosimetry.
Wisconsin: WILEY-VCH Verlag GmbH & Co. KGaA. Retrieved from
http://www-f9.ijs.si/~mm/Attix - Introduction to radiological physics and
radiation dosimetry.pdf
Bansal, G. J. (2006). Digital radiography. A comparison with modern conventional
imaging.
Postgraduate
Medical
Journal,
82(969),
425–428.
http://doi.org/10.1136/pgmj.2005.038448
Bath, M. (2010). Evaluating Imaging Systems: Practical Applications. Radiation
Protection Dosimetry, 139(1-3), 26–36. http://doi.org/10:1093/rpd/ncq007
Busch, H. P. (2006). Image quality and dose management in digital radiography : A
new paradigm for optimisation. Radiation Protection Dosimetry, (FEBRUARY
2005), 0–7. http://doi.org/10.1093/rpd/nci728
Colbeth. R., Boyce, S., Fong, R. (2001). 40×30 cm flat-panel imager for angiography,
R&F, and cone-beam CT application. Proc SPIE, 4320, 94-102
Commission, E. (1996a). European Guidelines on Quality Criteria for Diagnostic.
Luxembourg: Commission European.
Commission, E. (1996b). EUROPEAN GUIDELINES ON QUALITY CRITERIA FOR
DIAGNOSTIC. (J. H. E. (UK) Carmichael, C. (F) Maccia, B. M. (UK) Moores,
J. W. (D) Oestmann, H. (CEC) Schibilla, D. (CEC) Teunen, … B. (UK) Wall,
66
University of Ghana http://ugspace.ug.edu.gh
Eds.). Luxembourg: Commission, European.
Cygler, J. E., and Scalchi, P. (2009). MOSFET dosimetry in radiotherapy The authors
have received research support d Si. In AAPM Summer School.
Doi, K. (2006). Diagnostic imaging over the last 50 years : research and development
in medical imaging science and. Physics in Medicine and Biology, 51, 5–27.
http://doi.org/10.1088/0031-9155/51/13/R02
Donini, B., Rivetti, S., Lanconelli, N., and Bertolini, M. (2014). Free software for
performing physical analysis of systems for digital radiography and
mammography. Medical Physics, 41(5), 1–10.
Doyle, P. (2008). Assessment and optimisation of digital radiography systems for
clinical use. Assessment and Optimisation of Digital Radiography Systems for
Clinical Use. University of Glasgow. Retrieved from http://theses.gla.ac.uk/873/
Fahmi, M. M., Aisyahton, A., and Zaky, H. A. (2013). A Comparison of Two Indirect
Methods for Skin Dose In Chest Radiography. Health and the Environment, 4(2),
103–114. Retrieved from http://www.hej.kk.usm.my/pdf/
Freitas, M. B., and Yoshimura, E. M. (2009). Diagnostic reference levels for the most
frequent radiological examinations carried out in Brazil. Rev Panam Salud
Publica, 25(2), 95–104.
Herath, L., Chandrasiri, I. L. U., Senavirathne, S. A., and Rosairo, S. (2015). Entrance
Surface Dose Assessment for Postero-Anterior Erect Chest X-ray Examinations
of Adult Patients in a Selected Teaching Hospital in Sri Lanka, (November), 21–
25.
Hess, R., and Neitzel, U. (2012). Optimizing image quality and dose in digital
radiography of pediatric extremities. RoFo-Ortschritte Auf Dem Gebiet Der R,
184(7), 643–649. http://doi.org/10.1055/s-0032-1312727
Horizonte, B., Janeiro, R. De, Janeiro, R. De, Severiano, R. G., and Janeiro, R. De.
(2011). Assessment of dose in thyroid and salivary glands in dental radiology
using thermoluminiscent dosimetry. In International Nuclear Atlantic
Conference (pp. 1–8). Brazil.
IAEA. (2007). Dosimetry in Diagnostic Radiology: An International Code of Practice
(Technical). Vienna.
ICRP. (2006). Assessing Dose of the Representative Person for the Purpose of the
Radiation Protection of the Public. ICRP Publication 101a. Annals of ICRP (Vol.
36).
International
Commission
67
on
Radiological
Protection.
University of Ghana http://ugspace.ug.edu.gh
doi.org/10.1017/CBO9781107415324.004
Inkoom, S., Togobo, J., Emi-reynolds, G., Oddoye, A., Ntiri, T. O., and Gyekye, P. K.
(2012). Paper Retrospective Patient Dose Analysis Of Ghana ’ S First Direct
Digital Radiography System. Health Physics Society, 103(3), 133–137.
http://doi.org/10.1097/HP.0b013e3182475a76
Institute of Physics and Engineers in Medicine [IPEM] (2005). Recommended
Standards for Routine Performance Testing of Diagnostic X-ray Imaging
Systems, IPEM Report 91.
International Atomic Energy Agency (IAEA). (2013). HUMAN HEALTH SERIES
No. 24: Dosimetry in Diagnostic radiology for Paediatric Patients.
Vomfi.Univ.Kiev.Ua,
(24).
Retrieved
from
http://www-
pub.iaea.org/MTCD/publications/PDF/Pub1609_web.pdf\nhttp://www.vomfi.u
niv.kiev.ua/assets/files/IAEA/Pub1462_web.pdf
International Atomic Energy Agency (IAEA). (2014). Diagnostic Radiology Physics:
A Handbook for Teachers and Students. (D. R. Dance, S. Christofides, A. D. A.
Maidment, I. D. McLean, and K. H. Ng, Eds.). Vienna: International Atomic
Energy Agency (IAEA).
Jin Mo Goo, Im, J. G., Jong Hyo Kim, Joon Beom Seo, Tae Sung Kim, Sang June
Shine, and Lee, W. (2000). Digital chest radiography with a selenium-basecl flatpanel detector versus a storage phosphor system: Comparison of soft-copy
images. American Journal of Roentgenology, 175(4), 1013–1018.
Kandarakis, I., Cavouras, D., Panayiotakis, G. S. (1997). Evaluating x-ray detectors
for radiographic applications: a comparison of ZnSCdS:Ag with Gd2O2S:Tb and
Y2O2S:Tb screens. Phys Med Biol, 42, 1351-1373.
Körner, M., AU - Weber, C. H., AU - Wirth, S., AU - Pfeifer, K.-J., AU - Reiser, M.
F., and AU - Treitl, M. (2007). Advances in Digital Radiography : Physical.
Radiological
Society
of
North
America,
27(3),
675–686.
http://dx.doi.org/10.1148/rg.273065075
Kothan, S., and Tungjai, M. (2011). An Estimation of X-Radiation Output using
Mathematic Model Suchart Kothan and Montree Tungjai Department of
Radiologic Technology , Faculty of Associated Medical Sciences ,. American
Journal of Applied Sciences, 8(9), 923–926.
Kud, M. (2012). Image Quality Assessment. In WDS’12 Proceedings of Contributed
Papers, Part I (pp. 94–99).
68
University of Ghana http://ugspace.ug.edu.gh
Mattoon, J. S., and Smith, C. (2004). Breakthroughs in Radiography : LIMITATIONS
OF TRADITIONAL, (January), 1–8.
Mcparland, B. J. (1998). A study of patient radiation doses in interventional
radiological procedures. The British Journal of Radiology, 71, 175–185.
Mohammadi, P., Ebrahimi-moghadam, A., and Shirani, S. (2014). Subjective and
Objective Quality Assessment of Image : A Survey. Elsevier, (June), 1–50.
Mortazavi, S. M. J., Ghiassi-Nejad, M., Bakhshi, M., Jafari-Zadeh, M., Kavousi, A.,
Ahmadi, J., and Shareghi, A. (2004). Entrance Surface Dose Measurement on the
Thyroid Gland in Orthopantomography: The Need for Optimization. Iranian
Journal of Radiation Resources, 2(1), 1–7.
Neitzel, U., Maack, I., and Gunther-Kohfahl, S. (1994). Image quality of digital chest
radiography system based on a selenium detector. Med Phys, 21, 509-516.
Ovitt, T. W., Christenson, P. C., Fisher, H. D., Frost, M. M., Nudelman, S., Roehrig,
H., and Seeley, G. (1980). Intravenous angiography using digital video
subtraction: x-ray imaging system. AJR. American Journal of Roentgenology,
135(6), 1141–4. http://doi.org/10.2214/ajr.135.6.1141
Owoade, L. R., Sambo, I., and Tijani, S. A. (2015). Assessment of entrance surface
air kerma in patients undergoing chest X-ray from conventional diagnostic
radiology in Ogun State , Nigeria. International Journal of Medicine and Medical
Sciences, 7(6), 117–120. http://doi.org/10.5897/IJMMS2014.1119
Public Health England. (2016). Guidance: National Diagnostic Reference Levels
(NDRLs). Retrieved April 30, 2016, from
https://www.gov.uk/government/publications/diagnostic-radiology-nationaldiagnostic-reference-levels-ndrls/national-diagnostic-reference-levelsndrls#fnref:2
Puig, S. (2003). Digital radiography of the chest in paediatric patients in Germany.
Radiologe, 43, 1045-1050.
Radiation Products Design, I. (2014). Diodes, Dosimeters & Electrometers.
Albertville. Retrieved from www.rpdinc.com
Rehani, M. M. (2001). Protection of patients in general radiography. In International
Conference 26Y30 (p. 2001). Vienna: International Atomic Energy Agency.
Reis, C., Ndlovu, J., Serrenho, C., Akhtar, I., Haan, S. De, and Antonio, J. (2014).
Optimisation of paediatrics computed radiography for full spine curvature
measurements using a phantom : a pilot study. OPTIMAX 2014 - Radiation Dose
69
University of Ghana http://ugspace.ug.edu.gh
and Image Quality Optimisation in Medical Imaging, 43–51.
Reis, C. S. ., Ndlovu, J. ., Serrenho, C. ., Akhtar, I. ., Haan De, S. ., Garcia, J., …
Hogg, P. (2014). Optimisation of paediatrics computed radiography for full spine
curvature measurements using a phantom: a pilot study. Optimax 2014:
Radiation Dose and Image Quality Optimisation in Medical Imaging. Optimax
2014: Retrieved from http://usir.salford.ac.uk/34439/
Shahbazi-Gahrouei, D. (2006). Entrance surface dose measurements for routine Xray examinations in Chaharmahal and Bakhtiari hospitals. Iranian Journal of
Radiation Resources, 4(1), 29–33.
Song, X., Pogue, B. W., Jiang, S., Doyley, M. M., Dehghani, H., Tosteson, T. D., and
Paulsen, K. D. (2004). Automated region detection based on the contrast-to-noise
ratio in near-infrared tomography, 43(5), 1053–1062.
Sun, Z., Lin, C., Tyan, Y., and Ng, K. (2012). Optimization of chest radiographic
imaging parameters : a comparison of image quality and entrance skin dose for
digital chest. Journal of Clinical Imaging, 36(4), 279–286.
http://doi.org/10.1016/j.clinimag.2011.09.006
Taghi, M., Toossi, B., Akbari, F., and Roodi, S. B. (2012). Radiation Exposure to
Critical Organs in Panoramic Dental Examination. Acta Medica Iranica, 50(12),
809–813.
Tamboul, J. Y., Yousef, M., Mokhtar, K., Alfaki, A., and Sulieman, Abdelmoneim.
(2014). No Title. Life Science Journal, 11(2), 164–168.
Toivonen, M., Komppa, T., Authority, N. S., Ospedaliera, A., Della, S. M.,
Evangelismos, G. H., … Centre, R. (2003). Report on methods of evaluating local
skin dose in interventional radiology.
Tugwell, J., Everton, C., Kingma, A., Oomkens, D. M., Pereira, G. A., Pimentinha, D.
B., … Hogg, P. (2014). Increasing source to image distance for AP pelvis
imaging e Impact on radiation dose and image quality. Radiography, 20(4), 351–
355. http://doi.org/10.1016/j.radi.2014.05.012
Utsunomiya, S., Monzen, H., Akimoto, M., Mukumoto, N., Ishihara, Y., Shiinoki, T.,
… Hiraoka, M. M. . (2013). A feasibility study on reduction of the entrance
surface dose to neonates by use of a new digital mobile x-ray system.
Radiological Physics and Technology, 6(1), 157–161.
Vladimirov, A. (2010). Comparison of Image Quality Test Methods. Thesis Submitted
to the Faculty of Science and Technology Institute of Physics for an Award of
70
University of Ghana http://ugspace.ug.edu.gh
MSc Degree in Applied Physics (Medical Physics), 44.
Wanai, A., Kongkaew, P., and Yai, H. (2011). Entrance surface dose level of common
diagnostic x-ray examinations at Songklanagarind hospital. Songka Medical
Journal, 29(2), 57–64.
Wang, Z., Bovik, A. C., and Lu, L. (2002). Why is image quality assessment so
difficult? In Acoustics, Speech, and Signal Processing. Yorktown Heights, New
York.
WHO. (2016). Communicating Radiation Risks in Paediatric Imaging – InformatIon
to support healthcare dIscussIons about benefIt and rIsk. Geneva: World Health
Organization.
Williams, M. B., Krupinski, E. a., Strauss, K. J., Breeden, W. K., Rzeszotarski, M. S.,
Applegate, K., … Seibert, J. A. (2007). Digital Radiography Image Quality:
Image Acquisition. Journal of the American College of Radiology, 4(6), 371–
388. http://doi.org/10.1016/j.jacr.2007.02.002
Zhao, W., Rowland, J. A. (1995). X-ray imaging using amorphous selenium:
feasibility of flat-panel self-scanned detector for digital radiology, Med. Phys, 22,
1595-1604.
Zoetelief, J., Julius, H. W., and Christensen, P. (2000). Recommendations for patient
dosimetry in diagnostic radiology using TLD. IAEA-CN-96-46P, 92–93.
http://doi.org/10.1063/1.3047411
71
University of Ghana http://ugspace.ug.edu.gh
APPENDIX A
Table A1: Summary of individual patient exposures at Korle-Bu Teaching Hospital
Area of
Examination
Patient
Males Females
kVp
mAs
FFD
FSD
ESD
range
range
(cm)
(cm)
(mGy)
Thickness (cm)
Age (yrs)
Chest PA
17.2 - 23.8
18.0 - 81.0
33
72
62.0 - 68.0
12.5 - 18
180
156.2 - 162.8
0.12 - 0.71
Cervical spine AP
10.9 - 13.0
19.0 - 81.0
8
18
60.0 - 62.0
10.0 - 12.0
100
87.0 - 89.1
0.29 - 0.30
Cervical spine LAT
10.9 - 13.4
19.0 - 81.0
9
19
60.0 - 62.0
10.0 - 12.0
100
87.0 - 89.1
0.29 - 0.30
Lumbar spine AP
16.6 - 24.4
19.0 - 80.0
13
26
64.0 - 74.0
16.0 - 25.0
100
75.6 - 83.4
0.63 - 1.57
Lumbar spine LAT
20.5 - 26.5
21.0 - 80.0
17
26
80.0 - 95.0
36.0 - 40.0
100
73.2 - 79.5
2.46 - 4.19
Skull AP
18.0 - 20.4
19.0 - 63.0
7
12
72.0 - 75.0
22.0 - 32.0
100
79.6 - 82.0
1.19 - 1.87
Skull LAT
16.0 - 18.0
19.0 - 63.0
7
12
62.0 - 64.0
14.0 -16.0
100
82.0 - 84.0
0.52 - 0.54
Abdomen
17.1 - 23.7
19.0 - 72.0
9
22
73.0 - 83.0
25.0 - 45.0
100
76.3 - 82.9
1.27 - 3.49
72
University of Ghana http://ugspace.ug.edu.gh
Table A2: Summary of individual patient exposures at UGH
Area of
Examination
Patient
kVp
mAs
FFD
FSD
ESD
Thickness (cm)
Age (yrs)
Males
Females
range
range
(cm)
(cm)
(mGy)
Chest PA
11.0 - 27.0
18.0 - 81.0
49
53
125
2.5
180
153.0 - 169.0
0.30 - 0.36
Cervical spine AP
10.0 - 20.0
21.0 - 82.0
14
23
66.0 - 70.0
10.0 - 12.0
100
80.0 - 90.0
0.44 - 0.59
Cervical spine LAT
17.0 - 27.0
21.0 - 82.0
14
23
66.0 - 70.0
10.0 - 12.0
100
73.0 - 83.0
0.51 - 0.71
Lumbar spine AP
18.0 - 24.3
23.0 - 81.0
19
18
77.0 - 85.0
40.0 - 50.0
100
75.7 - 82.0
3.03 - 3.99
Lumbar spine LAT
21.0 - 31.5
23.0 - 81.0
19
18
96.0 - 98.0
40.0 - 50.0
100
68.5 - 79.0
4.50 - 7.12
Skull AP
18.0 - 21.5
19.0 - 73.0
8
11
75.0 - 78.0
32.0 - 34.0
100
88.5 - 92.0
1.65 - 1.79
Skull LAT
16.6 -18.0
19.0 - 73.0
8
11
70.0 - 74.0
16.0 - 18.0
100
92.0 - 93.4
0.71 - 0.74
Abdomen
17.5 - 24.4
24.0 - 74.0
9
12
77.0 - 85.0
40.0 - 50.0
100
85.6 -92.5
2.45 - 2.96
73
University of Ghana http://ugspace.ug.edu.gh
APPENDIX B
QC RESULTS AT KBTH
Table B.1: kVp accuracy results
Set voltage
Recorded
kVp
Exposure
Exposure time
(kV)
voltage (kV)
difference
(mGy)
(ms)
50
49.0407
-1.9186
0.0871
19.5880
60
59.2908
-1.1819
0.1474
19.5740
70
69.0489
-1.3587
0.2138
19.5926
80
78.7921
-1.5099
0.2871
19.5837
90
88.5723
-1.5864
0.3694
19.5833
100
98.5291
-1.4709
0.4598
20.0665
110
108.8694
-1.0278
0.5648
24.6019
120
119.3909
-0.5076
0.6720
25.1037
Table B. 2: kVp, Exposure and Timer reproducibility
Tube voltage
Exposure
Exposure time
(kV)
(mGy)
(ms)
78.7248
0.2856
20.0665
78.8805
0.2843
19.5829
78.8092
0.2855
20.0763
78.7437
0.2855
20.0763
78.8939
0.2843
20.0665
74
University of Ghana http://ugspace.ug.edu.gh
Table B. 3: mAs Linearity
Set mAs
Focal spot
Exposure (mGy)
mGy/mAs
0.50
Small
0.0069
0.0137
1.25
Small
0.0273
0.0219
2.20
Small
0.0553
0.0251
5.00
Small
0.1365
0.0273
11.00
Small
0.3159
0.0287
22.00
Small
0.6581
0.0299
40.00
Small
1.2289
0.0307
63.00
Large
1.9444
0.0309
80.00
Large
2.5308
0.0316
100.00
Large
3.2504
0.0325
Table B. 4: Timer Accuracy
Set time (ms)
Measured time (ms)
% Deviation
125
124.9780
-0.0176
80
79.8100
-0.2375
63
63.2450
0.3889
40
39.6570
-0.8575
32
31.5910
-1.2781
25
24.5650
-1.7400
20
20.0940
0.4700
16
15.5680
-2.7000
12
12.5390
4.4917
6.3
6.0480
-4.0000
75
University of Ghana http://ugspace.ug.edu.gh
Table B. 5: Tube Output measurement
Set voltage
Exposure
(kV)
Set mAs
(mGy)
Tube output (mGy/mAs)
40
50.0
0.2272
0.0045
50
40.0
0.3781
0.0095
60
32.0
0.4913
0.0154
70
20.0
0.4417
0.0221
90
16.0
0.5927
0.0370
109
10.0
0.5541
0.0554
117
8.0
0.5081
0.0635
125
6.3
0.4478
0.0711
141
4.0
0.2994
0.0748
150
2.0
0.1954
0.0977
76
University of Ghana http://ugspace.ug.edu.gh
APPENDIX C
QC RESULTS AT UGH
Table C. 1: kVp accuracy and Reproducibility results
Set kVp
Set mAs
Recorded voltage
kVp difference
(kV)
(mAs)
(kV)
(%)
60
32.0
64.45
7.4
60
26.0
64.18
7.0
70
32.0
73.69
5.3
70
16.0
73.56
5.1
70
8.0
73.29
4.7
81
32.0
88.06
8.7
81
16.0
87.85
8.5
81
8.0
87.08
7.5
Table C. 2: Timer Accuracy and Reproducibility
Set time
(ms)
49.2
Exposure time
(ms)
47.87
Time difference
(%)
-2.7
24.9
23.15
-7.0
49.2
48.84
-0.7
24.9
23.65
-5.0
12.9
11.58
-10.2
51.8
50.39
-2.7
25.9
24.68
-4.7
12.9
12.08
-6.3
77
University of Ghana http://ugspace.ug.edu.gh
Table C. 3: Tube Output measurement
kvp
mAs
Exposure (mGy)
Tube Output (mGy/mAs)
60
8
0.2395
0.0183
60
16
0.2634
0.0201
60
25
0.2713
0.0207
70
8
0.3305
0.0252
70
16
0.3664
0.0280
70
25
0.3739
0.0285
80
8
0.4502
0.0343
80
16
0.4909
0.0375
78
University of Ghana http://ugspace.ug.edu.gh
APPENDIX D
Table D.1: Datasheet used to collect exposure parameters at the hospitals
Area of examination
Chest AP/ Patient ID
1.
2.
3.
4.
5.
6.
7.
8.
9.
10
Lumbar spine AP/ Patient ID
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Patient
Age (yrs)
Sex
(M/F)
Thickness
(cm)
(V)
Tube
Current
mAs
Age
Sex
Thickness
kVp
mAs
79
kVp
FSD
ESAK
FFD
(cm)
mGy
(cm)
FSD
ESAK
FFD