The purpose of this study was to investigate image quality of a CT thorax examination with the question of issue being tumor obtained with a dose according to ALARA (As low as reasonably achievable).
Material and method
This study is an experimental study using a quantitative method. The data collection contained measurements on a Chest Phantom N1 (Kyoto Kagaku) inherited by Da Nang University of medical technology and pharmacy. This type of phantom is used to test or evaluate the image quality of a CT-scanner (11).
Chest Phantom N1
This phantom is designed for examinations with conventional x-ray and CT scanning. It allows us to perform images that are similar to the human body (chest). The phantom contains pulmonary vessels and bronchus up to the first bifurcation. It is placed in the phantom lung field. The arms of the phantom are placed so lateral and CT scanning is allowed. The phantom also has a mediastinum and pulmonary vessels that are put together as one piece (Figure 2 and 3).
In the mediastinum heart and trachea is included. The phantom consist of three main parts:
Main body (chest wall)
The size of the phantom is 45 cm and the measurements around the chest 94 cm. It weights 18 kg. In table 2 the material of the phantom is described.
Three types of simulated tumors were used. Table 3 and figure 4 shows the simulated tumors. They come in different color, Hounsfield number and material. Each simulated tumor is 10 mm in diameter and the shape sphere.
First the white simulated tumour (10mm) was put inside the phantom. The phantom was placed in the CT- scanner head first- supine. The first pace contained tests with different mAs to work out how many steps, were appropriate when reducing the mAs. The only parameter that was used was mAs because it is the best way of reducing the radiation dose (12). To come to a suitable reduction the tumour appearance was followed in each reduction of the mAs. This was made to see the differences in each image. With a small reduction of the mAs the image will look similar to each other, that’s why the big step of 15 was used.
This was done in consensus with the supervisors. After that an audit was made to analyse how much the image quality has been decreased in each mAs change. The conclusion was that four images for each simulated tumour were going to be performed with 15 steps (mAs steps) for each change. Except the last image, which only could be 13 because the machine could not go lower. This procedure was repeated for each simulated tumour. A total of 12 images were performed.
Protocol for ThorRoutine was used for imaging. First a Topogram of the phantom was performed. After that an image with the standard protocol was performed. For the second image the mAs/ref were lowered from 40/60 to 30/45. The third image was lowered even more from 30/45 to 22/30. The last image was lowered to the lowest possible, which is 17/17. Two different windows were used: mediastinal window and lung window. Ref which stands for mAs reference value is the pre-sets for the CT scanner. mAs stands for the value that the CT scanner gives automatically to the phantom.
This study is a blinded study meaning that the radiologists were anonymous and does not know where the tumors were put in the phantom. The writers did not know who the radiologists were. As a guide for the reviewing the European Guidelines for image quality (16) was used (figure 5). Appendix 3 shows a template for the radiologist to use when reviewing. Two radiologists were reviewing and sending their dictation back to us anonymous. The reason for this procedure is because the radiologists have different work experience and each of them must be able to review the images correctly with help of European Guidelines for image quality.
The images were named image 1, image 2 and image 3 so the radiologists did not know which tumors that were put in each image. Image 1 consisted the tumor with 100 Hu, image 2, the tumor with -800 Hu and image 3, consisted the tumor with – 630 Hu.
Statistical package for the social sciences (SPSS) has been used in this study for the calculations of how the mAs were increasing, which can be found in figure 6. Spearman correlation was used to calculate the correlation between the radiologist’s answers from the template and the correlation was considered significant at 0,01 level.
A radiographer works in a high technology environment with nursing, medicine, radiation physics, imaging and functional medicine. A radiographer operates for all human rights and culture rights, the right to live, the right to dignity and that everyone are treated with respect. Radiographers also conduct research and development and are responsible for research ethics guidelines (4).
Ethics is about building up, stimulate and keep alive a consciousness and a discussion about how to act. Science is important for both the individuals and for the development of society (17).
Since we are using a phantom no people are getting harmed. However, we have considered ethical issues in regard to Vietnamese culture and people, how the Vietnamese people live and work in Da Nang. The idea is to minimize the risk related to radiation for both staff and patients in conjunction with radiographic examinations. It is important for us to respect them and present our study without any bias. We are aware that our study might be used in the future and that we therefore have to be very steady. The advantage of this study is to decrease the radiation dose and increase the patient safety.
The results from the 2 radiologists reviewing shows that despite the decreased radiation dose, tumors were detected while maintaining a sufficient image quality. In the process of gathering research data, it was detected that the results from both of the radiologists were identical. Two of the questions in the European Guidelines were not relevant for this study. One of the questions was about contrast media, which were never used in the exams contained in this data collection. The other question was about if the lymphs could be visualized in the image but the phantom that has been used do not have lymphs. These two questions were the only ones that the two radiologists answered with no.
Table 4 shows the decrease of the radiation dose. kV 130 has been used for all the images and only the mAs has been changed with an decrease of 15 steps between each image acquisition. The last decrease could only be 13 because the CT-machine could not go any lower than that. The total radiation dose (DLP) has been decreased from 124 (mAs 60, which was the standard parameters) to 93 (mAs 45), 71 (mAs 30) and 58 (mAs 17). This is a decrease of 47 % (58/124=46,77*100=47%).
In the template for reviewing (appendix 3) the radiologists were asked to write a comment about what they found in the images, where the tumors could be found and in which image windows they could see it. They could find the tumors in every image but not all in both windows, often the tumors could only be found in the lung window. The two different windows are mediastinal (soft-tissue) and lung window. Mediastinal window is standard and lung window is a window with different HU then mediastinal window that help detect tumors easier.
Radiologist 1 could see four out of 16 tumors in both windows. All four tumors that could be seen belonged to image 1, which had the tumor +100 HU. Radiologist 2 had exactly the same comments on the reviewing, only the tumor with +100 HU could be seen in both windows.
Calculations using Spearman correlation showed that correlation coefficient was 1,00 between the radiologists’ answers. This means that the radiologists described the 16 images equally. The correlation is significant at 0,01 level.
The Hounsfield for lungs is approximately -800 HU. Water density has zero in HU and air has -1000 HU. The lungs are filled with air and are therefore near -1000 HU. The tumors used in this study had different HU and is visualized differently in the images. The closer the tumors HU are to -800 the harder they will be to find. One of the tumors has HU +100, which is a big difference from -800, therefore these tumors will be much easier to find. It does not matter how much the mAs is reduced these tumors will still be easy to detect (18).
That is the main reason why three different tumors were used, to do it more reality likely. Real tumors have different HU so just because one can be detected easy does not mean that another will be so simple to find, for example the tumor with HU -800. However, it is urgent that all tumors can be detected even with the lowest mAs for a sufficient assessment.
To be able to have a lower mAs as standard the images quality must be good enough to discover all types of tumors. Another reason for using three different tumors and use different location for the tumors in each image were in order to make it a little bit hard for the radiologists to find the tumors, also more reality likely. If the tumors were placed at the same spot in every image the radiologist would know where to look for the tumors and probably find it easier than if he did not know where it could be found.
The phantom is similar to a human body but it is in one shape and form. Real human bodies are more complex. The human body comes in different size, anatomy, age, gender and medical history. Some people may have been treated for different diseases and operations where the organs have been removed or medical devices have been implemented. Therefore, no human being are exactly identical to each other. The thorax phantom is simulating a standard healthy human body. Based on that, it may be easier to detect tumors in a phantom than in a real human being. This study could not be performed on real humans, since it is not ethical right to do that. Even if the tumors could be detected and diagnosed at mAs 17, the radiologist may not be able to detect at the same mAs on a real patient.
Although the mAs were decreased as much as possible the radiologists were still able to see the tumors in every image. Even if some tumors must have been harder to find they could still find them. It is always a risk to decrease the dose because the risk of missing something important increases. The radiologists were able to find the tumors at 17 mAs, which mean that there was a possibility to lower the dose to the patient and get a decent image quality for diagnosing. Previous research showed that by using mAs 25 the CT scan produced satisfactory image quality and also reduced the CTDI and in this way they were protecting patients from unnecessarily radiation. Our results and the results from this study indicate the fact that the radiation dose can be reduced and still keep a decent image quality (19).
Lower the mAs to 17 may be unreasonable, but make some kind of lowering seems fair. By studying the tables and how the mAs reduces the radiologists may found a reasonable lowering for the mAs, which will reduce the radiation doses to patients. When the radiologists reviewed the images they were able to visualize all the tumors using lung window. This shows the importance of using the lung window setting. In a study mentioned previously, results showed the percentage of normal-quality images was higher in the lung window than in the mediastinal window. In concordance with our study, this shows the need and the meaning of having the lung window setting (19).
A large amount of studies has been performed about CT radiation reduction. Even if there is a lot of information in the medical community about radiation reduction, not many professionals decide to use this information. As an example a typical low dose examination is stone protocol (for question kidney stone). It has been shown that more than 90% of these examinations were performed with higher dose technique. This study has shown that professionals have ignored recommendation of CT dose reduction. Our study showed that the radiation dose can be reduced, but it could still be a long way for this information to reach the actual optimization of the CT radiation dose in the radiology department (20).
Previous studies and the results from our study show that this information cannot be ignored for patient’s sake and therefor it is needed to take this information into the practical actual part of the optimization and also raise awareness of the importance of dose reduction (21). An example of how the information can reach the professionals would be that the physicians, radiologists and radiographers take this information into action. There is a need for a better cooperation between the professions. Each profession has the responsibility to constantly optimize the radiation dose and we as radiographer have a huge responsibility for the improvement (4). Sometimes it feels that we all are scared to try to reduce the radiation dose but this is an important topic that needs constant development.
Earlier studies have found an ability to lower the radiation dose with help from simple adjustments. For example: mAs and kV reduction, reduction of scan coverage and number of acquisitions. A phantom-study performed in China showed that they were able to reduce the effective dose by 53% to 90 % (20). In our study we were able to reduce the radiation dose with 47% and still get a decent image quality by only changing the parameter mAs.
ALARA is about optimization of CT radiation dose. It is not all about decreasing the radiation dose. It is about producing an acceptable diagnostic image with the lowest dose possible. Focus is image quality because that constrains the dose. Optimization is about doing the best for each individual. Dose optimization does not necessarily mean dose reduction because it can lead to poor image quality. An important factor is patient size. Different radiation doses need to be adjusted according to the patient size. Recommendation for radiation dose is 80 kV for infants, 100 kV for children, 120 kV for most adults, and higher voltages for obese adolescents and adults in CT examinations of abdomen and pelvis (21). In our study we used 130 kV and a phantom, which represent a normal sized patient.
Using these parameters on a much bigger patient would result in poor images quality. That is why it is important not to only focus on the radiation dose but the focus is on optimization of the CT radiation dose. Our result shows that the radiation dose could be lower while maintaining a good images quality on an average patient, but the lowering could probably not go that far if the patient was obese, because bigger patients require more radiation dose. We chose not to change the kV because it was a fixed parameter, we only changed the mAs but we think this is worth considering in every changing of parameters. Good images quality should be in focus and try to achieve that with the lowest radiation dose possible (21).
In this study a helical CT scanner was used. This type of scanner performs good images with decent quality and radiation dose. The fact that CT scanners are developing continuously makes the future bright for patients. An article about dual-energy of the lungs describes the current applications of Dual-energy CT, the imaging techniques in the thorax with focus on diagnosis and characterization of pulmonary disorders. They come to the conclusion that dual-energy CT can give both anatomic and functional information about the lungs. The quality of the dual-energy CT has shown improvement in diagnosing lung disorders (22).
It is important to keep developing the CT scanners for better diagnosing. Better technique, better quality, lower radiation dose might give us an opportunity to diagnose much faster and would make it easier for us as radiographers, for the radiologists and especially for the patients. The radiation dose has shown to be substantially lower than the reference value according to the European Guidelines on quality criteria for CT. By using the dual-energy technique it might obviate additional clinical tests (involving radiation investigations), which will decrease the net exposure to the individual patient. Today the main reason not using this type of technique might be the cost of it (22).
This study shows that is was possible to reduce the radiation dose while maintaining sufficient image quality when using a phantom with simulated tumours. The radiologists’ assessment was consistent which shows high credibility. Using the results the radiologist might come to the conclusion that a reduction of the mAs can be made. However, more research would be needed, and a suggestion could be to perform a similar study with real patients. This in order to evaluate the effect of the reduced radiation dose when detecting tumours in a human body of various sizes.
We want to say thank you to our supervisor in Sweden, Berit Björkman who has supported us through this thesis and journey. A big thank you to Dr. Dung which was our supervisor in Da Nang, Vietnam. We are very thankful for the cooperation with Danang University of Medical technology and pharmacy for letting us do the data collection at their student medical center.
Material and method
Chest Phantom N1
GET THE COMPLETE PROJECT
Investigation of radiation doses in conjunction with CT thorax examinations performed in Vietnam