Introduction
Radiation has been widely used in medical applications from diagnostic imaging to radiation therapy. In addition to its numerous advantages; however, the use of radiation carries risks of harmful effects on physical health. According to the 2008 report of the United Nations Scientific Committee on the Effects of Atomic Radiation, individuals have a cumulative medical radiation exposure of 0.6 mSv/yr worldwide, whereas the value is 2.6 mSv/yr in Japan [1]. A number of reports have discussed the risks of medical radiation exposure [2–10]. Most of these studies estimated cancer mortality and lifetime risk from radiation exposure via computed tomography (CT) scans in children or the rate of leukemia induction [11, 12]. Although the risk of adverse effects of radiation exposure in children is a major concern, it should not be ignored in adults. It has been estimated that 1.5%–2% of cancers in the USA are due to CT-based diagnostic procedures. While there is support for the linear no-threshold (LNT) hypothesis in International Commission on Radiological Protection (ICRP) Publication 103 ‘The 2007 recommendations of the International Commission on Radiological Protection’ [13], the LNT model (and its implications) have changed; it is not solely based on linear extrapolation to low doses. However, the Biologic Effects of Ionizing Radiation (BEIR) VII risk model is based on linear extrapolation, and many reports have made accurate estimates based on the model [14]. However, no conclusions have been reached regarding the pros and cons of adapting the data for atom bomb survivors to CT exposure.
Important exposure controls in health care include those for medical staff and patients. Three important points must be taken into account in the medical utilization of radiation: justification, optimization, and dose limits. However, dose limits are not established for patients when the potential benefit is greater than the exposure risk. Instead, the ICRP introduced the concept of diagnostic reference levels (DRLs) to optimize the protection of patients with regard to radiation exposure for diagnostic and interventional procedures as described in ICRP Publication 103 and ICRP Publication 105 ‘Radiation protection in medicine’ [15]. Medical exposure is classified as planned exposure, which includes caregivers and applications for biomedical research and is subject to the same dose constraints as occupational exposure, while DRLs apply only to patients.
DRLs are determined by collecting information on actual doses associated with imaging procedures for 20–30 patients of average build from many facilities, and then calculating median and third quartile (75th percentile) values. This dose should be defined as the dose index, i.e., a physical quantity that can be measured easily at each facility, rather than the effective dose (i.e., a ‘protective quantity’). The risk of stochastic effects is taken into consideration, while interventional radiology (IVR) also considers deterministic impact (i.e., tissue reactions). Imaging continues to evolve with advances in medical procedures and technology. Although the amount of radiation used at any one time is decreasing with advances in diagnostic imaging equipment and techniques, the number of examinations continues to increase, and the resultant increase in radiation dose to the population as a whole has become a social issue. In addition, radiation doses remain as high as ever at some facilities. However, DRLs are not dose limits or dose-constrained values; instead, they represent a standard for patient doses and can be used by facilities to optimize their doses. DRLs should be established by medical societies in collaboration with national health authorities and radiation regulatory authorities. Therefore, in addition to increasing awareness of DRLs, periodic reviews are necessary [16].
According to the Organization for Economic Co-operation and Development (OECD), Japan ranks first in the world for CT scanners, with 107 CTs per million people [17]. As Japan has the highest rate of CT use in the world, it also has the highest radiation dose. Among the DRLs for diagnostic radiation, this review will focus on CT imaging, discuss related issues, and analyze the trends and contents of previous studies. The characteristics of the 2015 and 2020 DRLs are compared based on changes in CT usage and new technologies over the past 5 years, and issues that have yet to be resolved are identified.
Materials and Methods
The status of exposure controls and DRLs in Japan is discussed based on the literature. Details of the 2020 DRLs set by working groups and project teams in which the author has participated are presented, along with issues to be addressed when setting future DRLs. For the sake of clarity, I will explain the historical matters of the last approximately 30 years in chronological order. The structure is as follows.
1. First medical exposure guidelines in Japan (1996–2005)
2. The ‘Lancet paper incident’ (2004)
3. J-RIME (2010)
4. 2015 DRLs in Japan (2013–2017)
5. 2020 DRLs in Japan (2017–2023)
6. Toward updating DRLs (2023–)
In principle, my research focused on literature presented in Google Search up to 30 pages of search results by the keyword DRL. In addition, I included guidelines related to DRLs that have been published by academic societies and organizations that are members of Japan Network for Research and Information on Medical Exposures (J-RIME) in Japan over the past 30 years. The references to the literature referred to during the formulation of DRLs 2020, in which I participated, are also added.
1. First Medical Exposure Guidelines in Japan (1996–2005)
The origins of DRLs in Japan are summarized by providing an overview of medical exposures and academic guidelines and medical exposure regulations. The first guidelines for diagnostic radiation doses in Japan, issued by the Japan Association of Radiological Technologists (JART) in 2000, are discussed based on the results of several questionnaires and a review of the literature.
2. The ‘Lancet Paper Incident’ (2004)
One of the major issues in raising awareness of medical exposure in Japan is the so-called Lancet paper incident in 2004. An overview of this case and its impact is provided below.
3. J-RIME (2010)
The establishment and operation of the J-RIME, which is currently leading medical exposure reduction efforts in Japan, is summarized.
4. 2015 DRLs in Japan (2013–2017)
On March 11, 2011, an accident at Fukushima Daiichi Nuclear Power Plant following the Great East Japan Earthquake caused widespread release of radioactive materials. Concerns about radiation in Japan, a country affected by atomic bombings at the end of World War II, intensified further, with some patients refusing even minimal radiation used in medical diagnostic procedures [18]. J-RIME established the 2015 DRLs at this time based on a nationwide questionnaire survey. DRLs for CT examinations were derived from the central and peripheral doses in acrylic phantoms 16 cm and 32 cm in diameter. The dose index was based on the computed tomography dose index volume (CTDIvol, in mGy) and the dose length product (DLP, in mGy·cm).
5. 2020 DRLs in Japan (2017–2023)
The status of activities of the CT Project Team (CT-PT) and Nuclear Medicine Project Team (NM-PT), in which the author participated, and the status of data collection and compilation for setting the 2020 DRLs are summarized. The procedure for setting the 2015 DRLs was followed, with DRL data collected through a nationwide survey and from relevant literature. Awareness-raising activities and ‘dissemination surveys’ were conducted during the 5-year period between 2015 and 2020. The 2020 DRLs were set based on how well they had penetrated the Japanese medical field and what problems had been encountered with the 2015 DRLs. ICRP Publication 135 ‘Diagnostic reference levels in medical imaging’ [19], issued in 2017, was also taken into consideration.
Results and Discussion
1. First Medical Exposure Guidelines in Japan (1996–2005)
As the first step toward establishing DRLs in Japan, JART published target values for reducing medical exposure in October 2000 in their ‘Medical exposure guidelines for patients (reduction target values)’ [20]. Based on the idea that radiological technologists are responsible for the medical radiation that they use in the course of their work, JART proposed target values for specific indicators based on international recommendations, such as ICRP Publication 73 ‘Radiological protection and safety in medicine’ [21] and International Atomic Energy Agency guidelines [22], as well as data from several domestic dose survey reports. CT examinations are important because the radiation dose is 10–100 times higher than for X-rays. Target values for CT examinations were defined based on a single-detector conventional scan system (Table 1). The dose is evaluated by multiplying the central dose of a cylindrical acrylic phantom 16 cm in diameter for the head and 30 cm in diameter for the abdomen by the absorbed dose conversion coefficient of acrylic. There are also a set of targets for CT fluoroscopy (Table 2), but there has been no mention of CT imaging in children.
Based on the Lancet paper incident in 2004, discussed below, the guidelines were revised in 2006 in ‘Dose reduction target value in radiological diagnosis: Medical Exposure Guidelines 2006’ [23]. When this revised version was issued, helical scanning with multiple rows of detectors had already become the norm; while this made it possible to perform imaging in a shorter time, the increase in exposure dose became a problem. The DRLs of the ICRP were adopted as international reference indices. Dose evaluation was also changed from absorbed dose to air kerma in the acrylic phantom, and computed tomography dose index weighted (CTDIw), which takes into account the central and peripheral regions of the phantom and the pitch of the helical scan CTDIvol, was adopted. Specifically, CTDIvol was obtained from May to December 2005 using 60 CT systems (16 single-slice CT systems and 44 multi-slice CT systems [2–64 rows]) at 50 facilities by reproducing the usual imaging conditions for phantoms. The average value of CTDIvol obtained from each facility was used as the reduction target value (Table 3). Pediatric CT guidelines were developed following the ‘Pediatric CT guidelines for reduction of radiation exposure’ published in February 2005 by the Japan Radiological Society (JRS) [24], Japanese Society of Radiological Technology (JSRT), and Japanese Society of Pediatric Radiology (Table 4). It was also mentioned that the irradiation conditions for CT fluoroscopy have changed significantly (Table 5).
2. The ‘Lancet Paper Incident’ (2004)
On February 10, 2004, a problem arose due to the content of a paper reported in a newspaper under the headline ‘Radiation exposure causes cancer, Japan is top of the world, British University Study.’ This paper, referred to as ‘the Lancet paper’ [25], caused a great deal of public concern as it reported that diagnostic X-rays could increase cancer incidence by 3.2% (7,587 cases per year) in Japan, in comparison to 0.6% in the UK and Poland, 0.9% in the USA, and 1.8% in Croatia. Diagnostic X-rays provide significant benefits. The level of exposure from diagnostic X-rays is usually low, and the individual cancer risk is very small. However, the probability of developing nine types of radiation-induced cancer (esophageal, stomach, colon, liver, lung, thyroid, breast, bladder, and leukemia) due to exposure to diagnostic X-rays up to age 75 years was estimated for the UK and 14 developed countries based on the LNT hypothesis. Public attention in Japan focused on the higher radiation exposure compared to other countries. Japanese researchers reported a number of statements to counter this bad publicity. Japan has the world’s highest per capita installation rate of CT machines, and CT is readily used because of its high diagnostic capability, which makes it possible to omit other X-ray examinations. Another reason for the high use rate of CT is the low cost of the examination due to Japan’s medical insurance system. When clearly stating the risk calculated by extrapolating the data of atom bomb survivors based on the LNT hypothesis to low doses, Japanese researchers stressed the benefits of radiation use, which is an essential consideration when discussing the problem of radiation exposure. Japanese researchers also noted that equating ‘no effect’ with ‘no detectable effect’ is problematic. Subsequently, Brenner et al. [2] evaluated cancer risk from exposure to CT scans in the USA. ICRP Publication 84 ‘Pregnancy and medical radiation’ [26], released in the same year, also had an impact. In Japan, the rapid shift to multi-row CT equipment was also a major challenge with respect to dose reduction. In 2007, the ICRP made a number of important recommendations regarding medical radiation exposure in Publication 103. These factors, associated with the Lancet paper incident, have increased interest in medical exposure in Japan, both among medical professionals and the general public.
3. J-RIME (2010)
Numerous issues related to medical exposure have been difficult for individual societies to address. In March 2010, J-RIME was established to collect data on facilities, equipment, dose frequency, exposure dose, and risk assessment in radiation practice, to understand the actual medical exposure situation in Japan and establish a medical exposure risk assessment and control system in Japan comparable to those in other developed countries. J-RIME has been working to address medical exposure issues throughout Japan, leveraging the strengths of government agencies, medical professionals, medical equipment manufacturers, and radiation protection experts. The organization started with 11 groups (chaired by Yoshiharu Yonekura in 2010–2017) and now consists of 20 groups (chaired by Makoto Hosono from 2017 to the present). At the time of the release of the 2020 DRLs, there were 17 member organizations, and the author participated in the DRL Working Group (DRL-WG) (Table 6).
4. 2015 DRLs in Japan (2013–2017)
J-RIME published ‘Diagnostic reference levels based on latest surveys in Japan: Japan DRLs 2015’ on June 7, 2015 [27]. The 2015 DRLs, the first DRLs in Japan to follow international standards, covered six areas: CT examinations, general radiography, mammography, intraoral radiography, IVR, and nuclear medicine.
The DRL values for CT examinations (adults) were established based on the results of two surveys. The first was a JRS survey (administered to 712 facilities) regarding the imaging conditions for all CT examinations performed on a single day from 19 to 25 May, 2014. We received responses from 443 facilities (797 CT and 24,860 examinations). The second survey used a questionnaire (administered by JART to 307 centers) requesting a description of typical imaging conditions for patients weighing around 65 kg by August 23, 2013. The DRL values for CT examinations (pediatric) were also established based on the results of two surveys. The first was a questionnaire on pediatric CT imaging conditions administered by JSRT to 339 facilities that had responded to a similar survey in the past; responses were received from 196 facilities by the deadline of August 31, 2012 [28]. The second survey (the JART questionnaire described above for adults) was used to obtain data for pediatric patients. For dose assessment, we used the CTDIvol (mGy) and DLP (mGy·cm) estimates displayed on the console of the CT system.
DRL values for adult and pediatric CT examinations are shown in Tables 7 and 8, respectively. Most protocols were based on a body weight of 50–60 kg. However, for coronary CT, the 75th percentile was obtained using 50–70 kg as the standard body size, with 60–70 kg used as the mode of measurement. For the JART survey, an adjustment was made to 55 kg, as 65±2.5 kg was considered the standard body size. For pediatric CT, two phantom sizes with diameters of 16 cm and 32 cm are listed because the phantom used for estimation may differ depending on the device.
In comparison with DRLs in other countries, Japanese doses were considerably higher for head examinations for both adults and children. The exposure doses for the chest and abdomen were slightly higher and lower, respectively, than in other countries.
The 2015 DRL values in nuclear medicine relate only to the radioactivity dose of the radiopharmaceutical administered and are undefined for the CT dose in single-photon emission computed tomography (SPECT)/CT and positron emission tomography (PET)/CT [29].
Furthermore, when the 2015 DRLs were published, DLP attracted attention as the DRL value for CT of the head was incorrectly stated as being more dangerous than living in a ‘no-go zone’ near the Fukushima Daiichi Nuclear Power Plant due to radioactive materials, with 1,350 mGy·cm exposure for the head erroneously reported as an enormous exposure dose of 1.35 Gy. As with the Lancet paper in 2004, the importance of correct reporting in the media was emphasized to academics.
5. 2020 DRLs in Japan (2017–2023)
J-RIME published ‘National diagnostic reference levels in Japan (2020): Japan DRLs 2020’ on July 3, 2020 [30], and coordinated the dissemination of information and research efforts by J-RIME member organizations. A number of activities were conducted from 2017 to 2020 to validate the 2015 DRLs and aid formulation of the 2020 DRLs [31–36]. J-RIME reorganized the DRL-WG to revise the DRLs in 2017. ICRP Publication 135 ‘Diagnostic reference levels in medical imaging’ [19] was published in 2017 as a global standard and reference for the next DRLs. The Japan Health Physics Society joined with J-RIME at this point, and the author participated in the DRL-WG. In 2018, the DRL-WG launched projects on seven additional areas of diagnostic fluoroscopy, including gastrointestinal angiography, endoscopy, and nerve block. The author joined the CT-PT and the NM-PT at this point.
In March 2019, the ‘Notification of the Director-General of the Medical Affairs Bureau of the Ministry of Health, Labour and Welfare No. 0312 No. 7,’ dated March 12, 2019, was issued as a partial amendment of the Ordinance for Enforcement of the Medical Care Act. The law came into effect in April 2000 and specified important considerations for the safe use of medical radiation. Radiation safety is now strictly regulated as part of the medical safety system, which includes a management system, hospital guidelines, staff training, medical exposure of patients, and explanations to patients. With regard to medical radiation exposure, recording and management of patient exposure doses became mandatory for CT, IVR, and nuclear medicine, and the ordinance clearly stated that the guidelines of related societies should be consulted. The 2015 DRLs played an important role in the guidelines of these related societies and were considered to be very important because they were published by J-RIME, which includes many Japanese organizations that deal with medical radiation. Efforts to accelerate the revision of DRLs for 2020 were undertaken, including incorporation into this statute.
The CT-PT, consisting of nine physicians and radiological technologists selected from J-RIME member organizations, was established in 2018. The CT-PT began by analyzing the 2015 DRLs to establish the 2020 DRLs and examined whether they were appropriate in light of the evolution of CT equipment, developments in medical technology, and changes in social conditions. Especially, the effects of cancer and other diseases caused by medical radiation exposure in children have been reported and are attracting attention, so a major revision to the DRLs was required. ICRP Publication 135 had a significant impact on the DRLs, and inclusion of the median value rather than only the 75th percentile was considered for optimization. The basic policy was to conduct a nationwide questionnaire survey to confirm the current status and make revisions as appropriate based on the recent literature.
The survey for adult CT examinations was conducted between September 30, 2019 and February 7, 2020, and 182 centers responded; the median number of consecutive procedures in patients aged 20–80 years of standard physique was 30 during any given period in 2019. Due to the small number of responses, we relied heavily on the report by Matsunaga et al. [31]. Standard body size was defined as body weight 50–70 kg for all protocols. Two new protocols were added: ‘acute pulmonary artery embolism and deep vein thrombosis’ and ‘trauma whole-body CT.’ The imaging protocols for acute pulmonary artery embolism and deep vein thrombosis varied among institutions, making standardization difficult. After repeated discussions within the CT-PT, the DRL values were established as a single imaging protocol based on the purpose of the examination, and not limited in terms of the examination site or imaging range. To survey trauma whole-body CT, questionnaire items were distributed among four protocols: head+trunk, head and neck+trunk, and head-to-trunk series. However, insufficient numbers of valid responses were obtained to set DRL values individually, so they were set with reference to the report by Miyayasu et al. [37].
Pediatric CT examinations were analyzed based on responses from 37 facilities to the ‘Survey of doses received by affected children in pediatric CT in Japan (2018)’ of the JSRT [38]. In addition, the results for 19 facilities of the Pediatric CT Dose Survey conducted from November to December 2018 by the Chiba Radiography Techniques Study Group were added. That survey was conducted between July 26, 2018 and February 29, 2020; survey forms were mailed to 409 facilities, including national, public, and private university hospitals, public medical institutions, and pediatric specialized medical institutions. The survey requested data on 50 consecutive cases for head, chest, and abdominal CT imaging protocols, and 37 facilities responded (valid response rate, 9%).
For dose assessment, the CTDIvol (mGy) and DLP (mGy·cm) dose outputs of the CT equipment were used as the DRL dose for adult CT, as in the 2015 DRLs. The values are based on a phantom 32 cm in diameter. The DRLs for pediatric CTs were also calculated in the same way as for adults. However, due to differences in the methods used by different CT systems to calculate dose estimates, values from a phantom 16 cm in diameter were used for head CT, and values from phantoms 16 cm and 32 cm in diameter were used for chest and abdominal CT.
DRL values for adult CTs are shown in Table 9. Unlike the 2015 DRLs, the standard body weights were all 50–70 kg. Two new protocols were then added. The published data also included the median dose distribution from the survey. In some protocols, results with imaging lengths less than one-half or more than twice the median were considered as data entry errors and thus excluded. The DLP values for the plain head routine were slightly increased relative to the 2015 DRLs, but those doses were left unchanged because there was no rational reason to increase it. The Japanese DRL values for plain head scans are 77 mGy for CTDIvol and 1,350 mGy·cm for DLP, respectively, which are higher (and have been for some time) than the values used overseas and require further optimization [39–53]. A comparison of a selection of representative countries is presented in Table 10. It was concluded that the increase in DRL values for liver scans was due to an increase in standard body weight. The number of facilities responding to the survey was 182, which was less than half the number that responded to the same survey at the time of the 2015 DRLs. Therefore, the data were likely to be affected by bias. The validity of combining trauma whole-body CT as one protocol must also be reconsidered.
DRL values for pediatric CT are shown in Tables 11 (by age) and 12 (by weight). In addition to age, the classification was also based on body weight in accordance with the guidelines outlined in ICRP Publication 135. For the chest and abdomen, values for the phantom 32 cm in diameter are shown in parentheses in addition to those of the phantom 16 cm in diameter. Among the aggregated results, the DLP values for the head and part of the abdomen were increased over the 2015 DRLs, but the dose was left unchanged as there was no rational reason to increase it. CTDIvol decreased from the 2015 DRLs at all sites. This was explained by a decrease in dose due to the replacement of CT systems with newer models with the capacity for ‘successive approximation reconstruction’ over the 5-year period from 2015 to 2020. Although the number of sites responding to the survey decreased significantly from 167 at the time of the 2015 DRLs, the median values did not differ significantly. The decrease in the number of responses was likely due to a number of factors, such as stricter ethics for clinical research, including surveys of adults. Since the revision of the Enforcement Regulations of the Medical Care Act has been clearly adopted as a guideline by the relevant academic societies, we expect that the next revision of the DRLs will increase the survey collection rate through requests for cooperation with data collection.
The NM-PT, consisting of nine physicians and radiological technologists selected from J-RIME member organizations, was launched in 2018. The NM-PT reviewed the 2015 DRLs for nuclear medicine examinations in adults to establish the 2020 DRLs, taking into account developments in medical technology and changes in social conditions. The NM-PT reviewed the 2015 DRLs for adult nuclear medicine testing, eliminating tests using radiopharmaceuticals that are no longer in use and adding new tests that are beginning to be performed. For children, following the 2015 DRLs, it was decided to defer to the Japanese Society of Nuclear Medicine ‘Consensus guidelines for appropriate practice of pediatric nuclear medicine examinations’ (revised 2019) (updated to version 2020 at the time of writing). As radiopharmaceuticals are also drugs, the dosage listed in the package insert was also observed, and care was taken to avoid discrepancies with the package insert and other documents while setting DRL values. As this paper refers to CT examinations, the details of DRL value setting for radiopharmaceuticals are left to the report by Abe et al. [54]. DRL values were added for hybrid CT devices used for absorption correction and image fusion, such as SPECT/CT and PET/CT. This was done in consultation with CT-PT but was consistent with the guidelines in ICRP Publication 135.
NM-PT sent a request to 915 medical facilities with nuclear medicine departments throughout Japan. The web-based survey (July 1–31, 2019) obtained the median actual dose for examinations of adults of standard body size and the median CTDIvol and DLP for hybrid imaging CT between August 26 and September 24, 2019. Valid responses were received from 256 facilities (28%), representing a significant decrease compared to the rate of 41% in the survey at the time of the 2015 DRLs. J-RIME concluded that ethics review was unnecessary because safety management of medical radiation has been legislated and implemented as a business practice. However, it is likely that more facilities have refrained from submitting data due to concerns over the violation of research ethics guidelines and the Personal Information Protection Law in relation to the submission of clinically relevant data outside the hospital. This is common to all imaging modalities and will be considered at the time of the next DRL revision.
DRL values for hybrid imaging CT used in nuclear medicine are shown in Tables 13 (SPECT/CT) and 14 (PET/CT). The CTDIvol and DLP of CT for SPECT/CT and PET/CT are shown for different purposes: ‘attenuation corrected only’ and ‘attenuation corrected+fusion image’ for SPECT/CT and ‘attenuation corrected+fusion image’ for PET/CT. Furthermore, to obtain images via hybrid imaging modalities equivalent to those used for diagnosis via CT alone, the examination should be performed in accordance with the DRL values of CT alone. Although the survey results showed a large degree of variation, it is difficult to evaluate the imaging equipment because detailed data regarding the equipment was not collected. However, as hybrid CT systems for nuclear medicine are very expensive and cannot be replaced frequently to keep up with the evolution of medical equipment, it was inferred that there would be large disparities among facilities. In addition, the variable current control and successive approximation reconstruction enabled by hybrid CT are not considered. Compared to similar reports from overseas [55], SPECT/CT and PET/CT were comparable except in terms of brain values, which were the same as those used in diagnostic CT and higher than those used overseas (Table 15). The DRL values for the brain were also higher when diagnostic and absorption correction alone were applied.
The main limitations when setting the DRL values for CT in the 2020 DRLs are discussed next. The number of facilities analyzed was 443 in the JRS survey and 439 in the study by Matsunaga et al. [31], and those results were referred to when setting the 2015 DRLs. The number of facilities in the present survey was smaller, at 182, and the degree of bias may therefore have been greater for ‘acute pulmonary artery embolism and deep vein thrombosis’ and ‘trauma whole-body CT.’ The data on pediatric CT may also have been biased due to the very small number of valid responses (9%). Similar trends were observed in the survey on nuclear medicine, where the proportion of valid responses dropped from 41% to 28%, and the results of some test types were unreliable. The provision of information to the center through legislation is an issue that remains to be resolved in the future.
Cone-beam CT in dentistry and angiography systems was not discussed by CT-PT or NM-PT as it is outside the jurisdiction of these project teams.
6. Toward Updating DRLs (2023–)
New technologies for CT systems, including phase-contrast CT, deep learning reconstruction technology, artificial intelligence reconstruction technology, photon counting detector system, Compton imaging technology, successive approximation reconstruction, and dual-energy imaging, are being applied clinically, and existing technologies are evolving [56–60]. Reports of reduced exposure using these methods have also begun to appear in the literature [61, 62]. Conversion from existing circular phantoms to size-specific dose estimates (SSDEs) has also been proposed. The use of SSDEs is not always effective for DRLs because they are not individual exposure studies, and further studies are required to optimize exposure reduction, improve image quality, and thus improve diagnostic performance. With regard to dose assessment, computer simulations of nonexposed areas involving the creation of a digital phantom for whole-body CT have been reported to estimate secondary cancer risk, etc. [63]. Accurate dosimetry and periodic audits are also necessary [64, 65].
It is important to improve the collection rate of questionnaires in nationwide surveys. DRLs must be collected as processed data in a manner that is ethical and does not violate the Act on the Protection of Personal Information, by making it well known that the data are collected for the revision of guidelines of related societies such that they comply with the Enforcement Regulations of the Medical Service Act. With the revision of the Enforcement Regulations of the Medical Service Act in 2020, the medical fee regulations have also been revised, and participation in dosimetry surveys is now mandatory for facilities for the purposes of image management. Although the additional fee for image management is only paid to large hospitals, such as those with specific functions, it is expected to be extended to all hospitals in Japan, for example, by transferring the obligation to participate in dosimetry surveys to the additional fee for image management. In accordance with the revised laws and regulations, dose control for CT, IVR, and nuclear medicine has become mandatory at all facilities, and data are collected in the original format, tabulated, and compared with DRLs to aid radiation control at each facility. Although not mandatory, radiation dose control is necessary for other imaging modalities, so some facilities record radiation doses for all modalities to the greatest extent possible. The Modality Performed Procedure Step (MPPS) of Digital Imaging and Communications in Medicine (DICOM), a medical imaging standard, as well as Radiation Dose Structured Reports (RDSR) [19] and systems that can automatically collect these data, are desirable.
Conclusion
Radiation is widely used in medicine and has both benefits and risks. Instead of radiation dose limits, DRLs are used to also take into account the benefits. DRLs are used to reduce radiation doses while ensuring adequate image quality and are not intended as strict limits. The ICRP recommends that they be reviewed periodically to keep pace with changes in social conditions and advances in medical technology and equipment.
This literature review discussed the historical organization of DRLs in Japan and the origins of DRLs in the field of CT imaging, presenting details regarding the setting of the 2020 DRLs by the project team of which the author was a member.
JART published Japan’s first medical exposure guidelines in 2000. J-RIME was founded in response to the Lancet paper incident and ICRP Publications 103 and 105, and the 2015 DRLs were set. Subsequently, the 2020 DRLs were set based on ICRP Publication 135 and revisions of laws and regulations in Japan. The DRLs for the brain for diagnostic CT, SPECT/CT, and PET/CT are higher in Japan than in Western countries, but the other DRLs are comparable between countries.
It will remain essential to periodically update DRLs according to changes in social conditions and advances in medical technologies and equipment. In particular, there has been remarkable progress in computing technology and dose reduction through successive approximation reconstruction methods. Deep learning and artificial intelligence reconstruction technology are also expected to become effective in reducing radiation exposure. New CT systems will greatly reduce exposure doses, which will have a significant impact on the setting of future DRLs. The collection of data regarding imaging conditions from many facilities in Japan will be necessary to facilitate future work regarding DRLs.
In addition, while only radiation exposure reduction is being evaluated at this time, it is expected that patient outcomes and clinical decision making associated with DRL updates in Japan will be evaluated in the future, given the history of their incorporation into medical safety.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation 
Print 
