MPM-D - Risk Assessment Centennial Ballroom 300D 14:30 - 16:45
|
| Chair(s): Darrell Fisher
|
| |
MPM-D.1
14:30 Machine Learning Methods and Multivariate Epidemiology in Radiation Risk Assessment Models H Lee*, Georgia Institute of Technology, Oak Ridge National Laboratory
; GA Agasthya, Oak Ridge National Laboratory; HA Hanson, Oak Ridge National Laboratory; JS Logan, Oak Ridge National Laboratory; JM Houri, Oak Ridge National Laboratory; AJ Kapadia, Oak Ridge National Laboratory; SA Dewji, Georgia Institute of Technology
Abstract: Radiation exposure is a known contributor to cancer risk. Unlike high dose exposure, the causal effects of low dose exposure are difficult to quantify because of difficulties in assessing cumulative exposures, long latency periods, and marginal increases in the overall cancer risk. In this analysis, machine learning (ML) was used to identify associations between county level rates of lung and bronchus cancer incidence and radiation exposure, while controlling for known confounders. For cancer risk, we used the 2013-17 age adjusted lung and bronchus incidence rate per 100,000 at the county level from CDC [avg: 63.4]. Radiation exposure was measured using airborne gamma counts, and radon concentration. We controlled for poverty, employment, education, percent urban, smoking, alcohol use, air quality, and age at the county level obtained from 3 different datasets. Linear regression (LR) and ML regression methods such as random forest (RF), gradient boosting (GB), and decision tree (DT) were used to estimate cancer risk. The ML methods showed higher R² values and lower root mean squared errors [R²: 0.778(RF), 0. 691(GB), 0.633(DT), RMSE: 6.96(RF), 8.26(GB), 8.98(DT)], compared to LR [R²: 0.587, RMSE: 9.56]. Causation effects were not obtained in this study due to the complexity of cancer development and the limits in data. However, a more sophisticated study design with diverse data of better quality can lead us closer to the causation in the future. Given ML’s ability to factor in the interactions between variables in the model, a necessity in complex phenotypes such as lung cancer, we expect that ML will provide a new perspective on the radiation epidemiology by reflecting the mutual influence of each contributor and modeling complex relationships. With progress in ML and curation of better quality and higher resolution spatio-temporal datasets, such epidemiology studies will lead to more accurate and detailed information on radiation exposure effects on human health.
|
MPM-D.2
14:45 How?The?Science of?Radiation Biology has?Helped?Remove the?Crippling?Fear of Low-Level Radiation??? AL Brooks*, Washington State University/Department of Energy (retired)
; JL Conca, UFA Ventures, Inc.; WM Glines, Department of Energy (retired); AE Waltar, Texas A&M University/American Nuclear Society (retired)
Abstract: Ever since the discovery of radiation there have been serious discussions on the beneficial and adverse health effects of low doses of radiation exposure. A series of early scientific uncertainties, along with the rise of many anti-nuclear movements, has caused a huge portion of the general public to become fearful of radiation at any level.?This fear or radiophobia was in part caused by the Linear No Threshold hypothesis, which states that each and every ionization increases the risk of cancer. ??This presentation reviews several other key scientific questions and hypotheses that arose in the early investigations of radiation?phenomena that generated early fears and, most importantly, the subsequent scientific findings that clearly demonstrate such fears have been exaggerated and are not supported by science.? The presentation will discuss the Linear No-Threshold Hypothesis, the risk for cancer, risk for genetic effects, the genetic load produced by radiation, the “hot particle” hypothesis and he role of 90 Sr on birth defects. The presentation will further articulate the costs that these unfounded concerns of low-level radiation (radiophobia) have caused in unnecessary and even hysterical responses to anything involved with the word “radiation.” The science focused on these early questions has overwhelmingly demonstrated a minimization of the health effects of low-level radiation.?Over the long haul it is critical to have Science, not Fear, win the battle to provide the huge benefits that radiation provides to our society.???
|
MPM-D.3
15:00 Dose Mapping Comparison Study Of Gamma Rays And X-Rays In Preclinical Models CS Gunther*, C&C Irradiator Service, LLC
; V Steri, University of California, San Francisco; JA Camara Serrano, University of California, San Francisco; J Caravaca Rodriguez, University of California, San Francisco; CV Nostrand, C&C Irradiator Service, LLC; Y Seo, University of California, San Francisco: University of California, Berkeley: Lawrence Berkeley National Laboratory
Abstract: The National Nuclear Security Administration’s Office of Radiological Security (NNSA/ORS) implements a radiological risk reduction program which seeks to minimize or eliminate the use of high activity radiological sources, including Cs-137, by replacing them with non-radioisotopic technologies. By working with volunteer sites, ORS is minimizing the risk of the theft and misuse of these sources. In pursuit of this goal, ORS is funding a number of research projects whose aim is to demonstrate the comparability of these alternative technologies by comparing results of gamma sources versus these technologies such as x-ray irradiators. The goal of this study, funded by ORS through Sandia National Laboratories, is to determine absorbed depth dose equivalence between gamma and x-ray technologies to remove any confounding factors in radiation experiments when the radiation source is switched between gamma and x-ray sources.
To this aim, we compared the absorbed dose rates from a Cs-137 gamma irradiator (J.L. Shepherd, Mark I) to an x-ray irradiator (Precision X-ray, X-Rad320) using the same dosimetry techniques on each device. We employed several dosimetry techniques: Ashland’s Dose-Map™ system, alanine pellets implanted inside rodent carcasses, and alanine pellets planted inside a 3D-printed Rodent BioPhantom (RBP).
Comparative analysis from each dosimetry technique resulted in an overall increase in dose versus the expected 25 Gy target throughout the x-ray energy spectrum (at 160, 225, and 320 kVp), and a decrease in dose versus the expected 25 Gy when using Cs-137 in three different depth locations of dose measurements (i.e., cranium, abdomen, and subcutaneous region). More precisely, the RBP and mouse carcass resulted in an increase of a 19.53% average dose versus the 25 Gy target when using the x-ray irradiator, while the RBP and mouse carcass resulted in a -7.47% decrease in average dose versus the 25 Gy target. From our datasets, the x-ray energies were more effective at producing absorbed dose throughout this study than the Cs-137 energy.
In conclusion, we generated dose mapping data and a dosimetry system which can be easily translated from gamma to x-ray irradiators, thus benefitting the biomedical research community as a whole and providing better animal welfare for in vivo irradiation practices.
|
MPM-D.5
15:15 CLONAL HEMATOPOIESIS OF INDETERMINATE POTENTIAL AND THE RISK OF EXPOSURE INDUCED DEATH FOR MARS MISSION SCENARIOS CM Werneth*, NASA Langley Research Center
; ZS Patel, KBR; SR Blattnig, NASA Langley Research Center; MS Thomposn, NASA Johnson Space Center; JM Pattarini, NASA Johnson Space Center; JL Huff, NASA Langley Research Center
Abstract: The space radiation environment is composed of ionizing particles that pose health risks to crew members embarking on deep space missions. NASA requires that astronaut career radiation exposures are limited such that cancer mortality shall not exceed 3% Risk of Exposure Induced Death (REID) at the 95% confidence level. The NASA Space Cancer Risk Model may be used to project the cancer REID for several radiosensitive organs and has been extended to estimate the REID for cardiovascular disease as well. A recent study by Jaiswal et al. (2014)† showed significant increases in incident coronary heart disease (CHD), ischemic stroke, and leukemia for individuals carrying gene mutations associated with clonal hematopoiesis of indeterminate potential (CHIP). CHIP is an age-related biological state observed in otherwise healthy individuals that is characterized by clonal expansion of hematopoietic stem cells carrying somatic mutations. Although CHIP can be found in all age groups, its prevalence increases with age and is found more commonly in individuals greater than 50 years old and in as many as 20% of individuals age 70 years and older. This study examines the increased REID associated with CHIP status for male and female astronauts during two Mars mission scenarios, one of which includes previous radiation exposure onboard the International Space Station.
(Jaiswal S, Fontanillas P, Flannick J, et al. Age-related clonal hematopoiesis associated with adverse outcomes.
N Engl J Med. 2014;371(26):2488-2498. doi:10.1056/NEJMoa1408617.)
*Disclaimer: This work was prepared while Z.S. Patel was employed at KBR/NASA JSC. The opinions expressed in this work are the author's own and do not reflect the view of the National Institutes of Health, the Department of Health and Human Services, or the United States government.
|
MPM-D.6
15:30 BREAK
|
MPM-D.7
16:00 RADIOLOGICAL ASSESSMENT OF COMMONLY CONSUMED FOOD CROPS GROWN IN RUSTENBURG, SOUTH AFRICA PO Olagbaju*, Physics Department, North West University, South Africa.
; OB Wojuola, Physics Department, North West University, South Africa.; VM Tshivhase, Centre for Applied Radiation Science and Technology, North West University, South Africa
Abstract: Ingestion of radiologically contaminated food crops is a source of internal radiation exposure to humans. This work reports radiological assessment of food crops grown in Rustenburg, one of the mining cities in South Africa for selected food crops: onion, beetroot, leeks, mints, parsley, maize and wheat. The measured average activity concentration of 238U in onion, beetroot, leeks, mints, parsley, maize and wheat are 0.51 Bq/kg, 1.43 Bq/kg, 0.25 Bq/kg, 0.26 Bq/kg, 0.13 Bq/kg, 25.28 Bq/kg and 0.54 Bq/kg respectively. The average activity concentration of 232Th was found to be 0.17 Bq/kg, 0.60 Bq/kg, 0.18 Bq/kg, 0.07 Bq/kg, 0.14 Bq/kg, 0.02 Bq/kg and 0.04 Bq/kg; while the average activity concentration of 40K is 25.52 Bq/kg, 37.35 Bq/kg, 36.01 Bq/kg, 38.36 Bq/kg, 45.11 Bq/kg, 10.60 Bq/kg, and 8.78 Bq/kg, in onion, beetroot, leeks, mints, parsley, maize, and wheat respectively. The total ingestion dose from the consumption of naturally-occurring radionuclides in onion, parsley, mint, maize, leek, wheat, and beetroot samples were estimated to be 94.13 µSv/yr, 162.98 µSv/yr, 138.94 µSv/yr, 290.42 µSv/yr, 130.86 µSv/yr, 75.80 µSv/yr and 141.39 µSv/yr respectively. The average committed effective dose in all investigated food crops were found to be below the reference dose of 120 µSv/yr for 238U and 232Th, 170 µSv/yr for 40K, and total dose of 290 µSv/yr reported by the United Nations Scientific Committee on the Effects of Atomic Radiation, except in cereals (maize and wheat). The estimated lifetime cancer risk from the consumption of food crops was also found below the threshold level of 2.90×10-04. Results obtained from the study indicates there’s no associated radiological risk with investigated food crops that were considered, making them safe for human consumption.
|
MPM-D.8
16:15 Should the Limit on Radiation Dose to the Public be Revised? DR Fisher*, Versant Medical Physics and Radiation Safety
Abstract: The public dose limit (1 mSv/y) recommended by ICRP and NCRP drives federal regulations on exposures from nuclear power plants, fuel cycle facilities, waste repositories, and some medical radiation sources. Before 1985, the public dose limit was set five times higher (5 mSv/y). To ensure compliance with the current ICRP public dose limit, the Environmental Protection Agency established an even more restrictive limit (0.25 mSv/y). And to comply with the EPA federal limit, site operational limits and remediation standards are usually set conservatively at a smaller fraction of the EPA limit, such as 1/10th to 1/3rd (0.025 to 0.08 mSv/y). But do ultraconservative dose limits make sense compared to natural background radiation (5 mSv/y) and the high costs of compliance? The 1 mSv/y public dose limit presumes that (1) future radiation-induced cancer incidence is predictable; (2) the risk of stochastic detriment from low-level, low-dose-rate radiation is known with high confidence; (3) the LNT dose-response hypothesis holds at annual dose levels between 0 to 20 mSv; and that (4) risk-modifying factors (dose rate, cellular repair, gender, age at exposure, smoking, diet) and other molecular-response modifiers are irrelevant and have no influence on underlying risk for all types of cancer. The consequence of progressively lower dose limits for the public is the higher cost of regulatory compliance. Measures needed to mitigate public exposures become exponentially greater as the dose limit is reduced; however, little or no tangible benefit may result in terms of avoided health effects. Did lowering the public dose limit from 5 mSv/y to 1 mSv/y yield any protective or beneficial value to society? The high cost of regulatory compliance is often prohibitive and unnecessary in terms of net gains in health and safety (cancer avoidance). With high assurance, it seems reasonable that members of the public, including the unborn embryo/fetus, may be adequately protected with a public dose limit of 5 to 20 mSv/y.
|
MPM-D.9
16:30 Panel
|
