Physics to understand biology: Monte Carlo approaches to investigate space radiation
doses and their effects on DNA and chromosomes
F. Ballarini1,2*, G. Battistoni2, F. Cerutti2,3, A. Ferrari2,4, E. Gadioli2,3, M.V. Garzelli2,3, A.
Mairani1,2, A. Ottolenghi1,2, L.S. Pinsky5, P.R. Sala2, S. Trovati1,2
1
University of Pavia, Nuclear and Theoretical Physics Department, via Bassi 6, Pavia, Italy
INFN (National Institute of Nuclear Physics), Italy
3
University of Milano, Physics Department, via Celoria 16, Milano, Italy
4
CERN, Geneva, Switzerland
5
University of Houston, Physics Department, Houston, TX
2
Abstract
Exposure of biological structures to ionizing radiation can induce different damage types at various levels, from
DNA and chromosomes up to cells, tissues, organs and entire organisms. Although these multi-step processes
involve many orders of magnitude both in the space and in the time scale, the pattern of initial energy deposition
in matter strongly influences the subsequent evolution of the process. Great help to elucidate the underlying
mechanisms and to perform reliable predictions is provided by mechanistic models and Monte Carlo codes,
which allow one to take into account the stochastic aspects characterizing energy deposition in matter.
Concerning radiation damage at the level of tissues and organs, in this paper we will focus on organ doses to
astronauts exposed to Galactic Cosmic Rays (GCR) in deep space, under different shielding conditions. The
calculations were carried out by means of the FLUKA transport and interaction MC code, coupled with two
anthropomorphic model phantoms inserted into an Al shielding box of variable thickness. Besides organaveraged absorbed doses and dose equivalents we calculated “biological doses”, defined as yields of clustered
DNA breaks (“Complex Lesions”) in a given organ. CL have been obtained by “event-by-event” radiation trackstructure simulations at the nm level and they are integrated on-line into a purposely modified version of
FLUKA, which adopts a “condensed-history” approach. To quantify the role of nuclear interactions, for each
shield thickness the dose contributions from secondary hadrons (including ions) were calculated separately.
Furthermore, the neutron contribution was separated from that of all other nuclear reaction products.
Concerning damage at the molecular and cellular level, herein we will present and discuss examples of
application of a Monte Carlo code developed at the University of Pavia, which can simulate chromosome
aberration induction by different radiation types. The focus will be both on the role played by the particle track
structure at the nm level and on the relationship between aberrations and relevant cellular endpoints such as cell
death and cell conversion to malignancy. Main assumption of the model is the hypothesis that only clustered
lesions (CLs) of the DNA double-helix can “evolve” and lead to aberrations. A combination of the two
approaches (condensed-history and event-by-event) allowed estimation of yields of chromosome aberrations
following exposure to GCR in deep space, which were found to be consistent with aberration yields observed in
lymphocytes of astronauts involved in long-term missions onboard the Mir station and the International Space
Station.
1. Introduction
As a natural continuation of the work presented at the 2003 Varenna conference
(Ballarini et al. 2003), where besides hadrontherapy we discussed calculations of organ doses
following exposure to Solar Particle Events (SPE) in deep space under different shielding
conditions, in this paper we will show analogous calculations for exposure to Galactic Cosmic
Rays (GCR). While the former are sporadic injections of charged particles – mainly protons
with energies below 200 MeV – coming from the Sun, GCR consist of protons, He ions and
*
Corresponding Author: Francesca Ballarini, University of Pavia, DFNT, via Bassi 6, I-27100 Pavia, Italy;
francesca.ballarini@pv.infn.it
1
heavier ions with energy spectra peaked around 1 GeV/n. The exposure to GCR, though
modulated by the Sun activity, is continuous. During the periods of minimum solar activity,
when the GCR flux is maximum due to the lower protection provided by the Sun magnetic
field, the GCR doses are of the order of 1 mSv per day, to be compared with the natural Earth
background which is about 1 mSv per year.
Absorbed dose (also called “physical dose”) and dose-equivalent, being average quantities,
are not sufficient to fully characterize the action of ionizing radiation, which is stochastic in
nature. Furthermore, radiation quality factors used to calculate dose equivalent were
introduced for radiation protection purposes. Therefore we have introduced a quantity, which
we call “biological dose”, defined as the yield of “Complex Lesions” per cell in a given organ
or tissue of interest. Complex Lesions (CLs), which are DNA breaks clustered within a few
tens DNA base-pairs, that is about 10 nm, have been calculated in a previous work by means
of “event-by-event” Monte Carlo track-structure simulations taking into account each single
energy deposition event - basically ionizations and excitations - at the nanometre level
(Ottolenghi et al. 1995). Since “event-by-event” approaches cannot be applied to large targets
such as organs or even entire organisms due to CPU time problems, we have developed an
approach consisting of on-line integration of CL yields into FLUKA (from FLUctuating
KAscades, since originally this code was used to predict calorimetric fluctuations on an eventby-event basis), which is a condensed-history Monte Carlo code which can simulate the
transport and interaction of electromagnetic and hadronic particles over a wide energy range in
a large set of materials (Aiginger et al. 2005). More details on the integration technique can be
found elsewhere (Ballarini et al. 2003).
The choice of Complex Lesions as a representative parameter for biological dose relies on
the fact that clustered DNA damage is widely recognized to be a fundamental step in the
processes leading to important damage types at sub-cellular and cellular level, such as
chromosome aberrations and clonogenic cell inactivation (Ottolenghi et al. 1997, Ballarini et
al. 2005). While cell inactivation consists of the loss of the cell ability to proliferate and form
colonies, aberrations are due to mis-repair of chromosome breaks leading to incorrect
rejoining of fragments belonging to different chromosomes, or to different regions of the same
chromosome. In particular in the last years we have been developing a mechanistic model and
a Monte Carlo code of chromosome aberration induction mainly based on the assumption that
aberrations arise from clustered DNA lesions. In section 3, after describing the main features
of the current version of the model and the code, we will present some preliminary
calculations aimed to estimate chromosome aberration yields following exposure to space
radiation.
2. Human exposure to Galactic Cosmis Rays: physical, equivalent and “biological” doses
2.1 Generalities on cosmic rays
A human mission to Mars, possibly preceded by the return to the Moon, is in NASA’s
plans within the first half of this century. Such mission would imply a long and continuous
exposure to Galactic Cosmic Rays (GCR) in deep space, that is without the protection of the
geomagnetic field. GCR fluxes consist of 87% protons, 12% He ions and 1% “HZE particles”
(i.e. particles with high charge and energy) originating from still unknown sources – maybe
supernovae explosions - outside our solar system. Although elements from Z=1 to Z=92 are
present, ions heavier than Iron are extremely rare. The energy spectra of the various GCR
2
components are peaked around 1 GeV/n, and the maximum energy values are of the order of 1
TeV/n. The intensity of Galactic Cosmic Rays is modulated by solar activity, which is based
on subsequent cycles lasting about 11 years. When the activity of the Sun is at its maximum,
the Sun magnetic field protects the inner solar system from the low-energy component of
GCR, whereas at solar minimum the GCR flux is approximately 2.5 higher (4 particles·cm-2·s1
) with respect to solar maximum. Although protons are the main contributors to the GCR
flux, heavier ions including HZE particles provide a substantial contribution to the dose
equivalent due to their higher Z and thus higher LET (Linear Energy Transfer, which is
proportional to Z2) and biological effectiveness (Ballarini et al. 2006, in press).
Exposure to GCR is therefore an important concern for long-term missions due to its
possible consequences for astronauts’ health, typically in terms of stochastic effects such as
cancer. Reliable risk estimations are mandatory in this field. However, the quantification of
space radiation risk is still affected by large uncertainties, which have been estimated to be in
the range 4-6 (Cucinotta et al. 2001). Such uncertainties are partly due to our poor knowledge
of the action of HZE particles particles, not only in terms of biological effects but also in terms
of nuclear interactions occurring in the spacecraft walls and shielding and in the human body
itself. Theoretical models and computer codes that simulate ion transport and interaction
including projectile and target fragmentation can be of great help. Examples of codes
developed by other groups and applied to space radiation related problems are HZETRN
(Wilson et al. 1995), SHIELD-HIT (Gudowska et al. 2004), PHITS (Iwase et al. 2002) and
HETC (Townsend 2005). While HZETRN is an analytical code, the others adopt a Monte
Carlo approach.
2.2 Methods and Results
In this section we will present calculations of GCR organ doses carried out with FLUKA.
In this context, it is worth mentioning that nucleus-nucleus interactions, which previously
were available only above 5 GeV/n, in the last few years have been implemented also for
lower energies. In the range 0.1-5 GeV/n the code presently adopts a modified version of a
Relativistic Quantum Molecular Dynamics approach (Aiginger et al. 2005; Ferrari et al., this
issue). In parallel, an original QMD code has been developed from scratch, and the interface
with FLUKA is in progress (Garzelli et al., this issue). Concerning energies lower than 100
MeV/n, it is also in progress an interface with a code based on the Boltzmann Master Equation
theory (Cerutti et al., this issue).
To allow for dose calculation in the various tissues and organs of the human body, the code
has been coupled with two anthropomorphic phantoms, a mathematical model (Pelliccioni and
Pillon 1996) and a "voxel" model (Zankl and Wittmann 2001). Both phantoms can be
inscribed into a shielding structure of variable shape, thickness and material. As in previous
studies (e.g. Ballarini et al. 2003, Ballarini et al. 2004), an Aluminum cylindrical shell was
used in the present work. The values selected for the box thickness were 1 g/cm2 (nominal
spacesuit), 2 g/cm2 (lightly shielded spacecraft) and 5 g/cm2 (nominal spacecraft). The space
between the box and the phantom was filled with air. Isotropic irradiation was performed from
an emitting sphere of 2 m radius, with GCR spectra taken from the model of Badhwar and O'
Neill (1996). All incoming ions with atomic numbers in the range 1-28 were considered.
As mentioned in the introduction, besides organ-averaged absorbed dose and dose
equivalent, "biological doses" were calculated as well, being the "biological dose" modeled as
the yield of clustered DNA breaks (“Complex Lesions" or CLs) per cell within a given organ
3
or tissue. For all calculations, the contributions from secondary hadrons were calculated
separately, in order to quantify the role of nuclear interactions with the shielding and with the
human body. Furthermore, the contributions from neutrons were separated with respect to
other (secondary) hadrons.
Figures 1a and 1b show, respectively, rates of RBM (Red Bone Marrow)-averaged
absorbed dose (in Gy/day) and “biological dose” (in (CLs/cell)/day) calculated with the
mathematical phantom as a function of the Al shielding thickness (in g/cm2). As expected
considering the high energy of the incoming particles, the total doses do not vary significantly
with increasing the shield thickness. Both for the physical and for the biological dose, the
contribution from nuclear interaction products is very similar to that of primary ions. While
the variation of the overall secondary-hadron contribution with increasing shielding is small,
the neutron contribution increases significantly; such increase is mainly determined by
neutrons originated in the shield rather than in the body. It is also worth mentioning that, at
least within the Al thickness range considered in these calculations, the neutron contribution
increases linearly, without showing saturations. Furthermore, the relative contribution from
neutrons is higher for the biological dose than for the physical dose, reflecting the high
biological effectiveness of these particles.
Figure 1a: RBM-averaged absorbed dose (in Gy/day) following GCR exposure in deep space at solar minimum,
as a function of the Al shield thickness. For each thickness, the different bars represent (from left to right): total
dose, primary-particle contribution, secondary-hadron contribution, neutron contribution, contribution from body
neutrons and contribution from shield neutrons (the latter obviously missing at zero shielding)
Figure 1b: RBM-averaged “biological dose” (as CLs/cell per day) following GCR exposure in deep space at
solar minimum, as a function of the Al shield thickness. For each thickness, the different bars represent (from left
to right): total dose, primary-particle contribution, secondary-hadron contribution, neutron contribution,
contribution from body neutrons and contribution from shield neutrons (the latter obviously missing at zero
shielding).
4
3. Chromosome aberrations following human exposure to cosmic rays
3.1 Generalities on chromosome aberrations
Cells exposed to ionising radiation can develop different types of chromosome aberrations
(CA) including dicentrics, translocations and complex exchanges, the latter usually defined as
rearrangements involving at least 3 breaks and 2 chromosomes. Chromosome aberrations,
nowadays observed by means of the FISH (Fluorescence In Situ Hybridization) technique,
which allows for selective painting of one or more specific chromosomes, are a particularly
relevant endpoint. Indeed aberrations represent an important step in the biological pathways
leading to both cell death and cell conversion to malignancy. While dicentric chromosomes
(i.e. chromosomes with two centromeres) cause a decreased probability for the cell to be able
to duplicate, some translocations (which consist of rearrangements where both the resulting,
aberrated chromosomes possess one centromere) are strongly correlated with cell conversion
to malignancy. In particular, a translocation involving the BCR and ABL genes on
chromosomes 9 and 22 is now regarded as causally related to the induction of Chronic
Myeloid Leukemia.
Chromosome aberration yields in human lymphocytes can also be used for biodosimetry in
case of accidents or exposure to complex radiation environments, typically mixed fields.
Monitoring of chromosome aberrations in astronauts' lymphocytes has become routine in the
last decade (Durante et al. 2003). While the aberration yields observed after short-term
missions generally are not significantly higher with respect to pre-flight levels, the crew
members of long-term missions tend to show increased levels of CAs. Comparison of postflight aberration yields with gamma-ray calibration curves can provide estimates of equivalent
doses, as well as of space radiation quality by taking into account measured absorbed doses.
Despite the significant advances in the experimental techniques and the large amount of
available data, some aspects of the mechanisms underlying the induction of chromosome
aberrations have not been fully elucidated yet. For example it is still a matter of debate
whether any DNA double-strand break (DSB) can participate in the formation of chromosome
aberrations, or if more severe – typically clustered - breaks are required. Theoretical models
and simulation codes can be of great help both as interpretative tools, for elucidating the
underlying mechanisms, and as predictive tools, for performing extrapolations where
experimental data are not available, typically at low doses and/or low dose rates. In the next
section we will present a model based on radiation track structure, which started being
developed various years ago (Ballarini et al. 1999). Furthermore, as a mere exercise, we will
show some calculations aimed to estimate chromosome aberration yields following exposure
to space radiation.
3.2 Methods and Results
The main assumption of the model (Ballarini and Ottolenghi 2003, Ballarini and Ottolenghi
2004, Ballarini and Ottolenghi 2005) consists of considering chromosome aberrations as the
“evolution” of clustered, and thus severe, initial DNA breaks, that is the Complex Lesions
mentioned above. This assumption relies on the fact that the dependence of CLs on radiation
quality reflects that shown by mutation and inactivation data (Belli et al. 1992), whereas nonclustered DSB show a much weaker dependence on the radiation type and energy. Each CL is
5
assumed to produce two independent chromosome free ends. Only free ends induced in
neighboring chromosomes or in the same chromosome are allowed to join and give rise to
aberrations, reflecting the experimental evidence that DNA repair takes place within the
channels separating the various chromosome “territories”, which are basically nonoverlapping intra-nuclear regions occupied by a single chromosome. The current version of
the model deals with human lymphocyte nuclei, which are modeled as 3-m radius spheres;
the 46 chromosome territories are described as (irregular) intra-nuclear domains with volume
proportional to the chromosome DNA content. Each territory consists of the union of small
adjacent cubic boxes. Repetition of chromosome territory construction with different
chromosome positions provides different configurations for lymphocyte nuclei in the G0 phase
of the cell cycle.
The yield of induced CL·Gy-1·cell-1 is the starting point for dose-response simulations.
While for photons the lesions are randomly distributed in the cell nucleus, for light ions they
are located along straight lines representing the cell nucleus traversals. The extension of the
model to heavy ions is in progress: as a first approach, a fraction of the lesions induced by a
heavy ion are “shifted” radially to model the effects of the so-called “delta rays”, which play a
significant role in determining the features of heavy particle tracks. For a given dose D (in
Gy), the average number of cell nucleus traversals n is calculated by n = Dr2/(0.16 LET)
where the LET is expressed in keV/m and r (in m) represents either the cell nucleus radius
(for light ions), or the nucleus radius plus the maximum range of delta rays (for heavier ions).
An actual number is extracted from a Poisson distribution. For each cell nucleus traversal,
random extraction of the point where the particle enters the nucleus provides the traversal
length, being the direction fixed (parallel irradiation). The average number of CLs per unit
length along a cell nucleus traversal is calculated as CL/m = 0.16 CL·Gy-1·cell-1·LET·V-1,
where V is the cell nucleus volume in m3. For each nucleus traversal, a Poisson distribution
provides an actual number of lesions.
Comparison of the CL positions to those of the boxes constituting the chromosome
territories allows association of the lesions to the various chromosomes. Specific background
(i.e prior to irradiation) yields for different aberration types (typically 0.001 for dicentrics and
0.005 for translocations) can be included. Both Giemsa staining (all chromosomes painted
with the same colour) and FISH selective painting can be simulated; small fragments, i.e. with
size of about 10 Mbp, are generally not scored since they can hardly be detected in
experiments. Simulation of CL induction and rejoining for a sufficiently high number of times
provides statistically significant aberration yields. Repetition of the process for different dose
values allows to obtain dose-response curves for the main aberration types, directly
comparable with experimental data. A schematic representation of the main simulation steps is
reported in Figure 2, whereas examples of simulated lymphocyte nuclei, each with its 46
chromosomes, are shown in Figure 3.
6
Dose

CLs/cell
(distributed according to track structure)

identification of hit chromosomes
free-end interaction
scoring
repetition for at least 100,000 cells
Figure 2: schematic representation of the main simulation steps of the chromosome aberration induction code
cell #1
cell #2
cell #3
Figure 3: examples of simulated lymphocyte nuclei, each with its 46 chromosome territories.
In previous works the model has been tested for gamma rays, protons and He ions by
comparing simulated dose-response curves with experimental data available in the literature,
without performing any fit a posteriori. The good agreement between model prediction and
experimental data for the induction of different aberration types allowed for model validation
regarding both the adopted assumptions and the simulation techniques. Furthermore, the
model has been applied to the induction of Chronc Myeloid Leukaemia (Ballarini and
Ottolenghi 2004) and to the estimate of dicentric chromosomes observed in lymphocytes of
astronauts following long-term missions onboard the Mir space station and the International
Space Station, on the basis of simulated gamma-ray dose response weighted by the space
radiation quality factor (Ballarini and Ottolenghi 2005). The extension of the model to heavy
7
ions has started only recently, and the results are still very preliminary. An example is reported
in Figure 4, which shows a dose response for simple exchanges (i.e. dicentrics plus
translocations) induced by 1 GeV/n Fe ions (LET=147 keV/micron). The line represents the
model prediction, whereas the points are experimental data taken from the literature (George
et al. 2003).
Figure 4: dose response for simple exchanges (i.e. dicentrics plus translocations) induced by 1 GeV/n Fe ions.
As a mere exercise, the two approaches described above (i.e. calculation of GCR Complex
Lesion yields by means of a modified version of FLUKA and calculation of chromosome
aberration yields starting from CLs) were combined, with the aim of providing an estimate of
chromosome aberrations induced by cosmic rays in deep space, starting from the Complex
Lesions calculated with FLUKA (0.0005 CL/cell per day). Given the GCR flux composition
(87% protons, 12% He ions and 1% heavier ions) and assuming as negligible the probability
for a cell nucleus to be traversed by more than one particle in deep space, CL yields by a
single cell nucleus traversal were calculated for 1 GeV/n protons, He ions and HZE particles.
Iron ions were taken as representative of heavy ions, which of course is a significant
approximation. Then, with the chromosome aberration induction code, the corresponding
chromosome aberration yields (expressed as aberrations/cell per day) were calculated.
While a single H or He traversal resulted not to give rise to aberration yields higher than
the background levels (due to their high velocity and thus low LET), a single Iron traversal of
the cell nucleus was found to induce chromosome aberration yields of the order of 0.26/cell
for dicentrics and translocations, and 0.45/cell for complex exchanges. Since a heavy-particle
traversal is a rare event, the aberration yields per cell per day are very small (of the order of
10-5). However, such yields become more important for long missions, e.g. about 0.01
translocations/cell following a deep-space mission of 600 days, which is the typical duration
of a possible mission to Mars. This number represents nothing else than the result of an
exercise; however, it is interesting that it is consistent with the highest post-flight aberration
yields observed in lymphocytes of astronauts involved in long-term missions onboard the Mir
space station and the International Space Station (Durante et al. 2003).
8
4. Conclusions
Monte Carlo approaches were applied to the simulation of space radiation organ doses and
chromosome aberration induction following exposure to ionizing radiation. More specifically,
organ-averaged absorbed doses and “biological doses” were calculated with the FLUKA code
for exposure to Galactic Cosmic Rays under different shielding conditions, resulting in doses
of the order of 0.5 mGy per day and 0.0005 CLs/cell per day, almost independent of shielding.
Furthermore, a track-structure based model and code of chromosome aberration induction was
presented, mainly relying on the assumption that only clustered DNA breaks can lead to
aberrations. Combination of the two approaches allowed estimation of chromosome aberration
yields following exposure to GCR in deep space. The resulting yields, mainly due to the heavy
ion component of the GCR spectra, were found to be consistent with post-flight
measurements performed on lymphocytes of astronauts employed in long-term missions
onboard the Mir station or the International Space Station.
Acknowledgements
This work was partially supported by the European Community (contract no. FI6R-CT-2003508842, "RISC-RAD").
References
Aiginger, H., Andersen, V., Ballarini, F., Battistoni, G., Campanella, M., Carboni, M., Cerutti, F., Empl, A.,
Enghardt, W., Fassò, A., Ferrari, A., Gadioli, E., Garzelli, M.V., Lee, K., Ottolenghi, A., Parodi, K., Pelliccioni,
M., Pinsky, L., Ranft, J., Roesler, S., Sala, P.R., Scannicchio, D., Smirnov, G., Sommerer, F., Wilson, T., Zapp,
N., The FLUKA code: new developments and application to 1 GeV/n iron beams. Adv. Space Res. 35, 214-222,
2005.
Badhwar, G.D., O'Neill P.M. GCR Model and its Applications. Adv. Space Sci. 17 (2), 7-17, 1996.
F. Ballarini, M. Merzagora, F. Monforti, M. Durante, G. Gialanella, G.F. Grossi, M. Pugliese, A. Ottolenghi
(1999), Chromosome aberrations induced by light ions: Monte carlo simulations based on a mechanistic model.
International Journal of Radiation Biology 75, 35-46.
Ballarini, F., Cerutti, F., De Biaggi, L., Ferrari, A., Ottolenghi, A., Parini, V. Importance of nuclear interactions
in hadrontherapy and space radiation protection, in: Gadioli, E. (Ed.), Proc. 10th Int. Conf. on Nuclear Reaction
Mechanisms, Varenna, Italy, June 9-13, 2003, pp. 635-643.
Ballarini, F., Ottolenghi, A. Chromosome aberrations as biomarkers of radiation exposure: modelling basic
mechanisms. Adv. Space Res. 31, 1557-1568, 2003.
Ballarini, F., Biaggi, M., De Biaggi, L., Ferrari, A., Ottolenghi, A., Panzarasa, A., Paretzke, H., Pelliccioni, M.,
Sala, P., Scannicchio, D., Zankl, M. Role of shielding in modulating the effects of solar particle events: Monte
Carlo calculation of absorbed dose and DNA complex lesions in different organs. Adv. Space Res. 34(6), 13381346, 2004.
Ballarini, F., Ottolenghi, A. A model of chromosome aberration induction and CML incidence at low doses.
Radiat. Environ. Biophys. 43, 2004.
9
Ballarini, F., Ottolenghi, A. A model of chromosome aberration induction: applications to space research.
Radiat. Res. 164, 567-70, 2005.
F. Ballarini, G. Battistoni, F. Cerutti, A. Ferrari, E. Gadioli, M.V. Garzelli, A. Ottolenghi, H.G. Paretzke, V.
Parini, M. Pelliccioni, L. Pinsky, P. R. Sala, M. Zankl, GCR and SPE organ doses in deep space with different
shielding: Monte Carlo simulations based on the FLUKA code coupled to anthropomorphic phantoms. Adv.
Space Res. 2006, in press.
M. Belli, D. T. Goodhead, F. Ianzini, G. Simone and M. A. Tabocchini, Direct comparison of biological
effectiveness of protons and alpha-particles of the same LET II Mutation induction at the HPRT locus in V79
cells. Int. J. Radiat. Biol. 61, 625-629 (1992).
Cucinotta, F.A., Schimmerling, W., Wilson, J., Peterson, L., Badhwar, G., Saganti, P., Dicello, J. Space radiation
cancer risk projections for exploration missions: uncertainty reduction and mitigation. NASA/JSC-29295,
January 2001
M. Durante, G. Snigiryova, E. Akaeva, A. Bogomazova, S. Druzhinin, B. Fedorenko, O. Greco, N. Novitskaya,
A. Rubanovich, V. Shevchenko, U. von Recklinghausen and G. Obe, Chromosome aberration dosimetry in
cosmonauts after single or multiple space flights. Cytogenet. Genome Res. 103, 40-46 (2003).
George, K., Durante, M., Willingham, V., Wu, H., Yang, T., Cucinotta, F.A., Biological effectiveness of
accelerated particles for the induction of chromosome damage measured in metaphase and interphase human
lymphocytes. Radiat Res 160, 425-435, 2003.
Gudowska, I., Sobolevsky, N., Andreo, P., Belkic, D., Brahme, A. Ion beam transport in tissue-like media using
the Monte Carlo code SHIELD-HIT. Phys. Med. Biol. 49, 1933-1958, 2004.
Iwase, H., Niita, K., Nakamura, T. J. Nucl. Sci. Technol. 39, 1142, 2002.
National Council on Radiation Protection and Measurements, Recommendations of Dose Limits for Low Earth
Orbit, NCRP Report 132, NCRP, Bethesda, M. D., 2000.
Ottolenghi, A., Merzagora, M., Tallone, L., Durante, M., Paretzke, H., Wilson, W. The quality of DNA doublestrand breaks: a Monte Carlo simulation of the end-structure of strand breaks produced by protons and alpha
particles. Radiat. Environ. Biophys. 34, 239-244, 1995.
Ottolenghi, A., Monforti, F., Merzagora, M. A Monte Carlo calculation of cell inactivation by light ions. Int. J.
Radiat. Biol. 72, 505-513, 1997.
Pelliccioni, M., Pillon, M., Comparison between anthropomorphic mathematical phantoms using MCNP and
FLUKA codes. Radiat. Prot. Dosim. 67, 253-256, 1996.
Townsend, L., NASA Space Radiation Transport Code Development Consortium. Radiat. Prot. Dosim. 116, 118122.
Wilson, J., Badavi, F., Cucinotta, F. et al., HZETRN: description of a free-space ion and nucleon transport and
shielding computer program. NASA Technical Paper 3495, May 1995.
Zankl, M., Wittmann, A., The adult male voxel model "Golem" segmented from whole-body CT patient data.
Radiat. Environ. Biophys. 40, 153-162, 2001.
10