This section, the results of the various measurements are presented based on the dose parameter estimate from dose report, dose assessment and risk assessment. Presentation of the summarized values of the experimental processes, including data analysis are presented below, in terms of the mean, median, the upper quartile (3rd quartile), maximum and minimum values based on age and gender variation of the measured parameters.
The head CT examination was the most frequently requested examination and constitute 50% of the total number, followed with abdomen-pelvis 30% and chest 20% as shown in Fig. 3.
Distribution of CT Examination by Age
The most dorminant age group was between 10–16 years and the least were 0–5 years which is understandable because CT is considered very high risk modality and are hence not recommended for neonates except in extreme cases. These are represented in Fig. 4
Estimate of Effective dose and Risk assessment
a) Using BEIR VII Method.
Table 4 and Table 6 shows the estimated of effective dose and risk assessment using the BEIR VII report based on age and gender variation.
Table 4
Effective dose and Risk assessment for male by BEIR VII method
Examination
|
Age
Years
|
Effective dose
|
Risk Incidence
|
Risk Mortality
|
SNR
|
mSv
|
%
|
%
|
|
Brain
|
0–5
|
1.3965
|
0.031620
|
0.01396
|
8.19
|
|
6–10
|
2.2785
|
0.039687
|
0.01877
|
12.08
|
|
11–16
|
4.5102
|
0.062799
|
0.03113
|
17.42
|
Chest
|
0–5
|
5.8213
|
0.140503
|
0.06110
|
5.11
|
|
5–10
|
15.889
|
0.276770
|
0.13093
|
9.53
|
|
11–16
|
24.059
|
0.335029
|
0.16606
|
16.30
|
Abdomen/pelvis
|
0–5
|
4.6980
|
0.113391
|
0.07918
|
8.07
|
|
6–10
|
12.334
|
0.214825
|
0.16014
|
9.32
|
|
11–16
|
13.965
|
0.187106
|
0.14356
|
15.11
|
b) Using ICRP method
Table 5 and Table 7 shows the estimated of effective dose and risk assessment using the ICRP publication for male and female.
Table 5
organ dose and Risk assessment for male by ICRP method
Examination
|
Age
|
Organ Dose
|
Risk Incidence
|
Risk Mortality
|
SNR
|
Years
|
mGy
|
%
|
%
|
|
Brain
|
0–5
|
10.83
|
0.15573
|
0.24501
|
8.19
|
|
6–10
|
11.18
|
0.16076
|
0.253
|
12.08
|
|
11–16
|
19.82
|
0.28501
|
0.44853
|
17.42
|
Chest
|
0–5
|
70.14
|
1.1345
|
1.14469
|
5.11
|
|
5–10
|
72.85
|
1.17871
|
1.18711
|
9.53
|
|
11–16
|
96.14
|
1.55541
|
1.56901
|
16.30
|
Abdomen/pelvis
|
0–5
|
0.06153
|
0.000163
|
0.000173
|
8.07
|
|
6–10
|
0.06319
|
0.000137
|
0.000137
|
9.32
|
|
11–16
|
0.00783
|
0.000171
|
0.00071
|
15.11
|
Table 6
Effective dose and risk assessment for female by BEIR VII method
Examination
|
Age
|
Effective dose
|
Risk Incidence
|
Risk Mortality
|
SNR
|
Years
|
mSv
|
%
|
%
|
|
Brain
|
0–5
|
1.374
|
0.057941
|
0.02199
|
6.03
|
|
6–10
|
3.19
|
0.102831
|
0.04142
|
6.85
|
|
11–16
|
3.846
|
0.096211
|
0.041
|
8.57
|
Chest
|
0–5
|
3.985
|
0.179196
|
0.06714
|
12.99
|
|
5–10
|
14.97
|
0.482555
|
0.19435
|
17.88
|
|
Nov-16
|
21.16
|
0.52923
|
0.22552
|
24.98
|
Abdomen/pelvis
|
0–5
|
3.516
|
0.158115
|
0.0369
|
12.7
|
|
6–10
|
11.36
|
0.366159
|
0.09359
|
15.72
|
|
11–16
|
13.74
|
0.328688
|
0.09184
|
25.71
|
Table 7
Organ dose and risk assessment for female by ICRP method
Examination
|
Age
|
Organ Dose
|
Risk Incidence
|
Risk Mortality
|
SNR
|
Years
|
mGy
|
%
|
%
|
|
Brain
|
0–5
|
10.98
|
0.15789
|
0.24848
|
6.03
|
|
6–10
|
15.65
|
0.22974
|
0.36516
|
6.85
|
|
11–16
|
15.65
|
0.22974
|
0.36516
|
8.57
|
Chest
|
0–5
|
64.96
|
1.04877
|
1.05955
|
12.99
|
|
6–10
|
68.41
|
1.10083
|
1.11571
|
17.88
|
|
11–16
|
92.6
|
1.49822
|
1.51038
|
24.98
|
Abdomen/pelvis
|
0–5
|
57.32
|
1.25072
|
1.62159
|
12.7
|
|
6–10
|
59.76
|
1.3039
|
2.1197
|
15.72
|
|
11–16
|
7.52
|
0.16373
|
0.21273
|
25.71
|
Analysis
Dose Optimisation
The basic framework of the presentation identifies the clear perspective of the relationship between the various parameters in tables representation and model equations, in addition to the Graphical User Interface [GUI]. For Head CT scan the tissues at risk are the brain and the lens of the eye. For chest CT scan the tissues at risk are esophagus, thyroid and lung. For the abdomen/pelvis CT scan the following organs; kidney, colon, liver, stomach and bladder at risk were considered.
The average values for organ dose and effective dose for Brain CT exam for age 0–5, 6–10 and 11–16 years were10.3mGy, 1.3965mSv; 11.18mGy, 2.2785mSv; and 19.82mGy, 4.5102mSv respectively with the corresponding image quality (SNR) as 8.19, 12.08 and 17.42 respectively for male paediatric patients.
Additionally, the average values for organ dose and effective dose for Chest CT exam for age 0–5, 6–10 and 11–16 years 70.1mGy,5.813mSv; 72.85mGy, 15.889mSv; 96.14mGy, 24.059mSv with corresponding signal to noise ratio for paediatric male patients are 5.11, 9.53 and 16.30 respectively.
Furthermore, the average values for organ dose and effective dose for Abdomen /Pelvis CT exam for age 0–5, 6–10 and 11–16 years were 61.5µGy, 4.698mSv; 63.19µGy, 13.33mSv; 7.83µGy, 13.965mSv and the corresponding SNR are 8.07, 9.32 and 15.11 respectively for male paediatric patients.
The details of the dose and estimated risk are captured in Tables 4 and Table 5 for male and female patients respectively.
The risk values were within the low range of 1 in 10,000 to 1 in 1,000 range as indicated in Tables 4 and 5 for male and female respectively.
Furthermore, the female aged 0–5 years estimated organ dose, effective dose and SNR were; 10.98µSv, 1.374mGy and 6.03 respectively. While the female aged 6–10 years estimated organ dose, effective dose and SNR were;15.65µSv, 3.846mSv and 6.85 respectively. Additionally, the organ, effective dose of aged 11–16 for female were 15.65µSv, 3.99mGy and 8.57 respectively. These age dependent increase trends were also observed in both the chest and abdominal pelvis examinations. However, it is important to point out that the estimated dose to the lungs was higher compared to both the brain and the kidney. These estimated values are similar as published in ICRP publication 121 [14]. Additionally, the quality of the images increased significantly as the dose increases in all the three regions (Head, Chest and Abdominal pelvis).
Furthermore, the ability to detect an aberrant object (lesion) in a radiograph is related to the proportion of the differential intensity to the ambient noise level. One important quantitative technique to determine this noise level is the absolute image signal to noise ratio. This ratio was measured on the CT images, the resolutions of which established a minimum SNR value of 8.2 with a corresponding minimum effective dose of 1.4mGy for head; 5.1 with a corresponding minimum effective dose of 5.8mGy for chest and a minimum value of 8.1 with a corresponding minimum effective dose of 4.7mGy for abdominal/Pelvis. This is significant as optimisation require a balance between the minimum image quality to the minimum dose to the paediatric patient. As this achieved in all the images.
Furthermore, for purposes of optimisation, it was important to establish the minimum dose level that could be used to answer the clinical question. That is, by obtaining the input parameters that could deliver the perfect images and at the same time that is diagnostically adequate for the specific health problem to be answered. This is the essence of optimization, which is to balance image quality with the corresponding radiation dose to the study tissues.
Additionally, the application of this optimization principle to CT imaging procedures requires a special approach, since too low a radiation dose could be as bad as a too high radiation dose which in both case the images obtained could be of unsuitable diagnostic quality. To achieve this a comprehensive Clinical Decision Support Application Software was designed to provide a user-friendly platform for comfortable working process. This is to allow the radiographers to predict the possible dose to the patients and with the reference chart the corresponding expected quality of image for the technique to use for effective optimization.
Estimating Of Cancer Risk Assessment
In estimating lifetime risks of cancer incidence and mortality, models are developed and used. These risk factors were estimated based on two approaches: Organ dose and risk assessment by ICRP method and Effective dose and risk assessment by BEIR VII method. This is because BEIR VII develops the most up-to-date and comprehensive risk estimates for cancer and other health effects from exposure to low-level ionizing radiation [14, 15]. These models take into account sex, age at exposure, dose rate and other factors [16]. Estimates are given for all solid cancers, leukemia and cancer of several specific sites. These risks’ models are based on data from Japanese atomic bomb survivors. Risk estimates are subject to several sources of uncertainty due to inherent limitations in epidemiology data. In addition to statistical uncertainty, the populations and exposures for which risk estimates are needed nearly always differ from those from whom epidemiology data are available [17].
Tables 5 and 6 gives the data set for Effective doses, organ dose, SNR and LAR values for all the examinations for cancer risk and mortality estimates.
In paediatric imaging optimisation is essential because it ensure that the produced image met the diagnostic request to enable all the questions to be answered. Where image noise does not distort the information for adequate and accurate analysis. This means the signal strength should not exceed the fivefold requirements that is recommended for adequate diagnostic decision. At the same time the dose to paediatric patients does not exceed the recommended values. Establishment of these procedure help to protect the patients from radiation dose which may result in a reduction of the risk to paediatric patient. This is because in an attempt to reduced dose to paediatric patients the quality of images to be produced is equally essential. That is reducing dose reduces the signal and thereby reduces the signal to noise ratio in the resulting CT image, hence, the image quality is affected. Hence it is necessary for to establish various diagnostic protocol to be followed before imaging.
Furthermore, the results of the model verification platform are presented in model equations. Additionally, the models are presented as risk assessment parameters, including modeled risk incidence and risk mortality. Finally, the input parameter which is strongly linked to output dose parameters were link to LAR in other to predict risk incidence and risk mortality presented as GUI for clinical application.
Hence, the models are presented as clinical application software for comfortable working process in dose optimisation. These were presented as dose parameter estimates software model, which has been designed for both exposures input and dose output data capturing mechanism. Both of these software models have been designed as GUI and Computer-aided design, CAD for use in clinical application. Hence it represents a graphical relationship between Modeled LAR, Input Parameters and Risk Incidence and mortality assessment. It also shows the model equations for both male and female risk assessments. In relation to incidence and mortality.
In conclusion, the modeled equations represent a linear relationship between LAR, Input Parameters and Risk Incidence. It is a linear approach for modeling the relationship between a scalar dependent, variables Risk Incidence and independent variables LAR and Input Parameters mAs and KVp.
Graphic User Interface (GUI) Model
Figure 5: shows the user interface for the estimate of risk incidence and mortality. It serves as a predictive model in diagnostic radiology.
Analysis Of Incidence And Mortality Risk Assessment
Table 8 shows risk level and the additional risk factors of fatal cancer from CT imaging. While Table 9, Table 10 and Table 11 are Modelled equations for Head, Chest and Abdomen/Pelvis respectively.
Table 8
Risk of fatal Cancer from CT Examination
Risk Level
|
Approximate additional Risk of fatal Cancer from CT Examination
|
Negligible
|
Less than 1 in 1,000,000
|
Minimal
|
1 in 1,000,000 to 1 in 100,000
|
Very Low
|
1 in 100,000 to 1 in 10,000
|
Low
|
1 in 10,000 to 1 in 1,000
|
Moderate
|
1 in 1,000 to 1 in 500
|
Furthermore, there is an urgent need for most the radiology departments in Ghana to improve the radiation protection issues in the departments since it appeared that all the regional sections studied exceeded the IAEA/ICRP recommended effective dose estimates. However, the signal to noise ratio over exceeded the minimum values of five, which is adequate to answer the clinical questions. Twelve incidence and mortality risk modeled equations were obtained summarized below:
Table 9
Head Model
|
Model Equation
|
Male Incidence
|
Y = 0.0501–0.000030 LAR + 0.000131X1 + 0.000268X2
|
Female Incidence
|
Y = 0.1098–0.000033 LAR + 0.000004 X1 + 0.000517 X2
|
Male Mortality
|
Y’ = 0.0548–0.000062 LAR - 0.000000 X’1 + 0.000108 X’2
|
Female Mortality
|
Y’ = 0.0718–0.000052 LAR + 0.000001 X’1 + 0.000184 X’2
|
Table 10
Chest Model
|
Model Equation
|
Male Incidence
|
Y = 0.228–0.000079 LAR - 0.00056 X1 + 0.00137 X2
|
Female Incidence
|
Y = 1.379–0.000203 LAR - 0.00276 X1 - 0.00209 X2
|
Male Mortality
|
Y’ = 0.488–0.000344 LAR - 0.000730 X’1 - 0.000679 X’2
|
Female Mortality
|
Y’ = 0.689–0.000295 LAR - 0.00113 X’1 - 0.00095 X’2
|
Table 11
Abdomen/Pelvis model equations
Abdominal-Pelvis Model
|
Model Equation
|
Male Incidence
|
Y = -0.239 + 0.000083 LAR - 0.00013 X1 + 0.00183 X2
|
Female Incidence
|
Y = -0.438 + 0.000082 LAR - 0.00018 X1 + 0.00337 X2
|
Male Mortality
|
Y’ = -0.1203 + 0.000094 LAR + 0.000668 X’1+ 0.000595 X’1
|
Female Mortality
|
Y’ = -0.246 + 0.000110 LAR + 0.000719 X’1+ 0.001358 X’1
|
Where,
Y is cancer risk incidence
Y’ is cancer risk mortality
X1 is the mAs
X2 is the kVp
LAR is the life time attributable risk