Introduction
Chest radiography is one of the most frequently used screening procedures among various radiological examinations. In addition to screening, chest radiography may be performed on consecutive days to evaluate changes in the lungs, heart, and other organs of the same patient. Although patient exposure in radiography is relatively low compared to other radiological examinations, it has been reported that the number of examinations is large compared to other modalities [1]. Therefore, it is important to understand dose management and patient exposure during chest radiography.
The diagnostic reference level (DRL), proposed by the International Commission on Radiological Protection (ICRP), is occasionally used for dose management [2]. Many countries and regions have reported DRLs, and entrance-surface air kerma (ESAK) and air kerma-area product (KAP) are used as dose indicators in radiography [3–7]. These doses can be easily measured and used to assess whether doses are optimized at each facility. However, the doses used in DRLs cannot be used to evaluate the actual dose to which the patient is exposed. To evaluate the patient dose, organ-absorbed dose and effective dose need to be calculated. The organ-absorbed dose shows the absorbed dose at the location of the target organ in the human body. The effective dose can be estimated by multiplying the organ-absorbed dose by the radiation weighting factor and tissue weighting factor specified by the ICRP and adding them to the total [8]. The organ-absorbed dose is the relevant quantity for defining organ- and tissue-specific dose limits associated with tissue reactions, whereas the effective dose serves as a protection quantity for comparing radiation exposures across modalities and for communicating dose-related information to patients [9, 10]. In radiography, these doses can be used to explain that the patient exposure is lower than that from other modalities.
Organ-absorbed doses and effective dose can be estimated either by inserting a dosimeter into a human equivalent phantom and measuring the actual dose or by using simulations or conversion coefficients. Simulations and conversion coefficients provide a means of estimating organ-absorbed and effective doses without the need for dedicated dosimeters. Kelaranta et al. [11] estimated the organ-absorbed dose conversion coefficients for the breast and lungs in chest posterior–anterior (PA) radiography from incident air kerma, while Chen et al. [12] and Ono and Asada [13] estimated the effective dose conversion coefficients from the KAP. In particular, Chen et al. [12] derived conversion coefficients for each pelvic radiography at different angles of incidence, and these values differed from each other. Most studies on conversion coefficients for chest radiography have been based on the PA direction, and few reports have examined differences in the incidence direction. The breast, for which the tissue weighting factor is set high by the ICRP [8], is located in the anterior part of the human body, so the organ-absorbed dose is expected to be higher in the anterior–posterior (AP) than in the PA. Similarly, in lateral radiography, the organ-absorbed dose, effective dose, and effective dose conversion coefficient may differ between left lateral (Lt Lat) and right lateral (Rt Lat) radiography due to the difference in the arrangement of organs in the body, but the extent of this difference has not been clarified.
The aim of this study was to evaluate the effect of different incidence directions on the organ-absorbed dose, effective dose, and effective dose conversion coefficients for chest radiography. In particular, we aimed to clarify the extent to which AP and PA radiography differ and that to which Lt Lat and Rt Lat radiography differ.
Materials and Methods
1. Setting up a Monte Carlo simulation
In this study, the organ-absorbed doses and effective dose were estimated using the Monte Carlo simulation software PC Program for X-ray Monte Carlo (PCXMC version 2.0; STUK). PCXMC is a dose estimation software used in radiography and interventional radiology and is available for six different age ranges, from 0 year old children to adults [14, 15]. It has also been reported to have a higher accuracy for organ-absorbed doses than that with other simulation software [16].
Fig. 1A shows an example of a PCXMC simulation screen. Information such as the subject’s age group, height, and weight can be entered; however, the sex of the subject cannot be selected. An examination environment for radiography can be constructed by inputting the irradiation field size and source-to-detector distance (SDD) and setting an arbitrary X-ray incidence point. The number of photons in the simulation reached about 2 billion. Increasing the number of photons improves the calculation accuracy; however, it also increases the computational time. Fig. 1B shows an example of the calculated results. By entering the incident air kerma (mGy) or dose-area product (mGy·cm2), the organ-absorbed dose and effective dose can be obtained according to the simulation model. The organ-absorbed dose can be estimated not only for organs with tissue weighting factors but also for various organs and tissues in the body.
Table 1 lists the simulation conditions used in the study. As phantom data, we assumed a PCXMC standard body size (178.6 cm in height and 73.2 kg in weight). The number of photons was set uniformly with 500 million, and the irradiation field size was kept constant at 35.4 cm×43.0 cm at the detector surface position to reduce the variation of the irradiation area between the examinations. In clinical practice, the distance between the detector and the patient varies depending on the patient positioning. In PA and lateral radiography, the detector is often positioned within the detector stand, whereas in AP radiography, it is commonly placed directly behind the patient. Accordingly, the phantom exit-to-image distance was adjusted according to the incidence direction.
2. Estimation and Evaluation of Organ-Absorbed Dose
The simulation conditions are listed in Table 2. The tube voltage was set to 85 kVp for AP, 120 kVp for PA, and 140 kVp for Lt Lat and Rt Lat, as described in previous studies [3, 5, 17]. ESAK was set as 0.3 mGy for frontal radiography and 1.0 mGy for lateral radiography, with the incident air kerma excluding backscatter as the input dose to PCXMC. Frontal radiography doses were based on Japan’s DRLs. Lateral radiography doses were estimated from the frontal values using dose ratios between frontal and lateral radiography reported in various countries [5, 6, 18]. KAP was calculated by converting the irradiation field size in Table 1 to the area of the incident surface location and multiplying it by the input dose. Based on the corresponding simulated environment and input doses, the organ-absorbed doses were estimated for frontal radiography (AP and PA) and lateral radiography (Lt Lat and Rt Lat).
The organ-absorbed doses were classified into organs within the chest region and organs in the extrathoracic region, and the magnitude of the doses was evaluated for each organ. This classification was devised independently by the authors. Therefore, the organs and tissues within the chest region, as specified by PCXMC, were selected as the breasts, extrathoracic airways, oesophagus, heart, lungs, thymus, thyroid, skin, active bone, and skeleton. The remaining organs were selected as the brain, oral mucosa, salivary glands, stomach, liver, gall bladder, pancreas, spleen, kidneys, adrenals, colon (large intestine), small intestine, urinary bladder, prostate, testicles, uterus, ovaries, lymph nodes, and muscles, which were set as organs and tissues in the extrathoracic region. In PCXMC, skeletons can be estimated separately as skulls, spines, and ribs, but in this study, only doses that are displayed as ‘skeletons’ were evaluated. Similarly, for the ‘colon (large intestine),’ separate absorbed doses for the upper large intestine and the lower large intestine can be estimated, but in this study, the ‘colon (large intestine)’ dose was used for evaluation.
3. Estimation and Evaluation of Effective Dose and Effective Dose Conversion Coefficient
The effective dose was estimated using the conditions in Table 2. Similar to the estimation of organ-absorbed doses, effective doses were estimated for AP and PA and Lt Lat and Rt Lat, respectively. The conversion coefficients were calculated using Equation (1), according to previous studies [13].
The relative difference (%) of the effective dose with respect to PA or Lt Lat radiography was calculated using Equation (2).
Effective doseAP or PA represents the effective dose of AP or PA radiography; for lateral radiography, AP was replaced by Rt Lat and PA by Lt Lat in Equation (2). The effects of different incident directions on the effective dose and effective dose conversion coefficients were evaluated by comparing the obtained doses and relative differences.
Results
1. Evaluation of Organ-Absorbed Dose
Fig. 2A shows the organ-absorbed dose in the chest region, while Fig. 2B shows that in the extrathoracic region for frontal radiography. For the breast and thymus, the difference in the organ-absorbed dose between AP and PA radiography was more than 5-fold. The organ-absorbed doses by the heart and thyroid were also higher in the AP group than those in the PA group. Conversely, the lungs and skeletons were larger in PA than in AP. For organs in the extrathoracic region, the largest difference in absorbed dose was observed in the adrenals, where the PA dose was approximately 6.8 times higher than the AP dose. In contrast, dose differences among the other extrathoracic organs were relatively small.
Fig. 3A shows the organ-absorbed dose in the chest region, while Fig. 3B shows that in the extrathoracic region for lateral radiography. For organs in the chest region, the heart was approximately 1.5 times larger in Rt Lat than in Lt Lat; however, there were no organs where the difference was several times larger than that observed with frontal radiography. Regarding organs in the extrathoracic region, the stomach, pancreas, and spleen had higher absorbed doses in Rt Lat than in Lt Lat, and the liver had a higher dose in Lt Lat than in Rt Lat. In particular, a difference of approximately 15.5 times was observed in the spleen.
2. Evaluation of Effective Dose and Effective Dose Conversion Coefficient
Table 3 shows the effective dose, relative difference, and effective dose conversion coefficient. In frontal radiography, the effective dose was approximately 30% higher for AP radiography than for PA radiography. However, the difference in the effective dose for lateral radiography was not as large as that for frontal radiography. The effective dose conversion coefficient for frontal radiography was 40% higher for AP radiography than for PA radiography. For lateral radiography, the error was equivalent to the effective dose, with an Rt Lat value that was approximately 6% higher.
Discussion
Diagnostic X-rays can be used to find diseases within the body but may slightly increase the risk of developing cancer [19, 20]. Therefore, it is very important to be aware of the various patient exposures during radiological examinations. However, these doses represent doses to reference humans, which differ slightly from the doses to individually exposed patients [10]. In particular, the tissue weighting factors use uniform values regardless of sex or age, and the location and risk of developing cancer vary depending on the sex and age at the time of exposure. It should be noted that when these doses are used to communicate with patients, they may not be the exact dose or risk received by each individual patient.
We used PCXMC to calculate the organ-absorbed doses and effective doses, which represent the doses in a phantom previously determined within a Monte Carlo simulation. Although the doses correspond to the phantom used in the simulation, which differs from the human body, this study provides a computational approach that is useful for approximate estimation. Previous studies have reported errors ranging from a few percent to 20% between the doses obtained using the Rando phantom and PCXMC [13, 16]. There are also several types of reference phantoms published by the ICRP, including voxel phantoms [21] and adult mesh-type reference computational phantoms [22], and the recommended reference phantoms are likely to change in the future. Therefore, the organ-absorbed doses and effective doses obtained in this study have some errors compared with the original doses. However, because the arrangement of the organs in the phantom does not change significantly, there is no significant difference in the rate of change in the dose due to differences in the incidence direction.
Regarding the organ-absorbed doses, the breasts and thymus are located on the anterior side of the body, so the absorbed doses in the AP radiography are considered to be higher. Conversely, the lungs are anatomically elongated along the posterior aspect of the chest, and since organs such as the heart are positioned anteriorly, the absorbed dose to the lungs tends to be greater in PA radiography. Furthermore, although skeletal structures are distributed throughout the chest region, components such as the vertebrae and scapulae are located posteriorly and are therefore thought to exhibit a similar tendency. The dose to organs in the extrathoracic region near the diaphragm, such as the stomach and liver, was higher than that to other intra-abdominal organs because of the scattered X-rays from organs inside the irradiation field and a small amount of direct X-rays. For lateral radiography, the difference in the absorbed dose in the stomach due to differences in the direction of incidence was similar to that reported by Matsunaga et al. [23]. For other organs in the extrathoracic region, the spleen on the left side of the abdominal region showed a larger difference in the organ-absorbed dose due to the difference in the direction of incidence. In addition, the liver showed a larger absorbed dose in the Lt Lat radiography, and the dose did not differ from that of the organs in the irradiation field. In this study, the irradiation field was set at the position of the detector surface. Because the longer the SDD, the wider the irradiation field on the incident point side, we considered that the organs near the diaphragm were exposed to X-rays in the irradiation field. In the chest region, the different incident direction of the X-ray beam resulted in a large difference in organ-absorbed doses during frontal radiography, while lateral radiography exhibited little variation. In contrast, in the abdominal region, there was no large difference in organ-absorbed doses in frontal radiography (with the exception of some organs), and a large difference in organ-absorbed doses was observed in lateral radiography. These results suggest that the organ-absorbed doses may be reduced by appropriately selecting the direction of incidence during radiography.
When estimating the effective dose, the organ-absorbed dose is multiplied by the tissue weighting factor determined in ICRP Publication 103 [8]. This factor varies widely among organs/tissues, with the bone marrow (red), colon, lung, stomach, and breast having the highest values of 0.12. Regarding Table 3, we considered that the effective dose was higher for AP and Rt Lat radiography because the breast is located on the front side of the body and the stomach is located on the left side of the body. For reference, the difference in effective dose according to differences in incident direction exhibited a trend similar to findings from previous studies in the spinal region [24]. The effective dose conversion coefficient for frontal radiography was approximately 40% higher for AP radiography than that for PA radiography. In this study, the SDD was set at 120 cm in the AP direction only, and the irradiation field size was constant for AP and PA radiography. Therefore, the irradiation field area at the patient’s body surface position was larger for PA radiography, resulting in a larger KAP for PA radiography. However, the effective dose per KAP was higher for AP radiography because the effective dose was higher for AP radiography. Therefore, we considered that the ratio of the difference was increased by the conversion coefficient rather than by the effective dose. The effective dose conversion coefficients can easily calculate the effective dose; however, the errors can be large if care is not taken regarding the direction of the incident radiation.
In this study, based on previous studies, SDD and tube voltage were varied according to the incident direction in the frontal projection. Differences in tube voltage can affect organ-absorbed doses and effective dose, as X-ray attenuation within the body varies depending on energy. Even with the same incident air kerma, lower tube voltage may result in higher doses at the body surface. Additionally, differences in SDD lead to variations in the irradiated field at the body surface, which may influence both KAP and the effective dose conversion coefficients. These factors represent limitations of the study. Nevertheless, the data obtained under conditions resembling clinical practice may serve as a useful reference for clinical applications.
Conclusion
We evaluated the effects of different incident directions on organ-absorbed doses, effective doses, and effective dose conversion coefficients for chest radiography. The organ-absorbed doses of some organs/tissues differed greatly depending on the direction of incidence. For organs/tissues in the chest region, the difference in absorbed dose tended to be larger on frontal radiography, whereas in the abdomen region, the difference in absorbed dose tended to be larger on lateral radiography. The effective dose was approximately 30% higher for AP than for PA radiography. The effective dose conversion coefficient was approximately 40% higher for frontal AP radiography. Our results are not surprising given the arrangement of organs in the body, but they have not been specifically reported before. It was suggested that using the PA direction of incidence in frontal radiography reduces the effective dose received by the patient. In addition, when estimating the effective dose using conversion coefficients, the estimated value may differ significantly due to differences in the direction of incidence, so the coefficients must be appropriately selected.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation 
Print 
