Journal of Trace Elements in Medicine and Biology 31 (2015) 204–208

Contents lists available at ScienceDirect

Journal of Trace Elements in Medicine and Biology

journal homepage: www.elsevier.com/locate/jtemb

10th NTES Symposium

A neutron activation technique for manganese measurements

in humans
C. Bhatia ∗ , S.H. Byun, D.R. Chettle, M.J. Inskip, W.V. Prestwich
Department of Medical Physics and Applied Radiation Sciences, McMaster University, Hamilton, ON L8S 4K1, Canada

a r t i c l e i n f o a b s t r a c t

Article history: Manganese (Mn) is an essential element for humans, animals, and plants and is required for growth,
Received 30 January 2014 development, and maintenance of health. Studies show that Mn metabolism is similar to that of iron,
Accepted 28 July 2014 therefore, increased Mn levels in humans could interfere with the absorption of dietary iron leading
to anemia. Also, excess exposure to Mn dust, leads to nervous system disorders similar to Parkinson’s
Keywords: disease. Higher exposure to Mn is essentially related to industrial pollution. Thus, there is a benefit
Occupational and industrial exposure
in developing a clean non-invasive technique for monitoring such increased levels of Mn in order to
Manganese toxicity
understand the risk of disease and development of appropriate treatments.
Accelerator neutron source
Non-invasive technique
To this end, the feasibility of Mn measurements with their minimum detection limits (MDL) has been
Neutron activation reported earlier from the McMaster group. This work presents improvement to Mn assessment using an
upgraded system and optimized times of irradiation and counting for induced gamma activity of Mn. The
technique utilizes the high proton current Tandetron accelerator producing neutrons via the 7 Li(p,n)7 Be
reaction at McMaster University and an array of nine NaI (Tl) detectors in a 4␲ geometry for delayed
counting of gamma rays. The neutron irradiation of a set of phantoms was performed with protocols
having different proton energy, current and time of irradiation. The improved MDLs estimated using
the upgraded set up and constrained timings are reported as 0.67 ␮gMn/gCa for 2.3 MeV protons and
0.71 ␮gMn/gCa for 2.0 MeV protons. These are a factor of about 2.3 times better than previous measure-
ments done at McMaster University using the in vivo set-up. Also, because of lower dose-equivalent and a
relatively close MDL, the combination of: 2.0 MeV; 300 ␮A; 3 min protocol is recommended as compared
to 2.3 MeV; 400 ␮A; 45 s protocol for further measurements of Mn in vivo.
© 2014 Elsevier GmbH. All rights reserved.

Introduction
Manganese is a trace element that is an essential nutrient for
normal development and body function across the life span of all
living organisms [1]. In humans, the diet is the main source of
Mn, with an adequate intake of 2–4 mg/day and an upper toler-
able intake of 11 mg/day for total Mn intake [2]. The major storage
of Mn is in the bones (about 50%) and its excretion is through the
liver [3]. A healthy person with normal liver and kidney function
can excrete excess dietary manganese. Inhaled Mn is of greater con-
cern because it can bypass the body’s normal defense mechanisms;
i.e. it is often transported directly to the brain via olfactory nerve
[4] before it is metabolized by the liver. Also, inhalation exposure to
high concentrations of manganese dusts can cause an inflammatory
∗ Corresponding author. Tel.: +1 (905) 525 9140x26328.
E-mail addresses: bhatiac@mcmaster.ca, chitraphy@gmail.com (C. Bhatia).
response in the lung, which, over time, can result in impaired lung
function.
Manganese is one of the most widely used metals in industry
[5]. Exposure to Mn dust occurs primarily in mining, and metal-
lurgical operations for iron, steel, ferrous and nonferrous alloys.
Manufacturing of dry-cell batteries, anti-knock gasoline additives,
pesticides, pigments for ceramics, dyes, and matches can also lead
to occupational Mn exposure. Manganese fumes are produced
during metallurgical operations and several types of welding oper-
ations. The principal sources of Mn in the atmosphere are however,
from natural processes including volcanic gas and dust, and forest
fires [6].
Exposure to Mn is usually via inhalation, which can lead to Mn
accumulation and results in adverse health effects including dam-
age to the lungs, liver, kidney and central nervous system [7,8].
Mn toxicity has been reported through occupational (e.g. welder,
miner) and dietary over exposure. Prolonged exposure to high
Mn concentrations (>1 mg/m3 ) in air may lead to a Parkinsonian
syndrome known as “manganism” [9–12]. Numerous studies have

http://dx.doi.org/10.1016/j.jtemb.2014.07.018
0946-672X/© 2014 Elsevier GmbH. All rights reserved.
 C. Bhatia et al. / Journal of Trace Elements in Medicine and Biology 31 (2015) 204–208
indicated that welders may be at high risk of Parkinson’s disease
and neurological health effects [9–12].
The majority of the available scientific literature on Mn expo-
sure, body burden and its health effects comes from animal studies
and invasive methods like biopsies and autopsies. Toxic elements
like Mn are commonly measured in blood and serum, or by a bone
biopsy. It is widely accepted that blood and serum concentrations
of elements provide information only about the recent exposure to
the element, and therefore do not necessarily reflect chronic expo-
sure of an individual. On the other hand, bone biopsy is painful,
involves a risk to the patient and it may not be possible to repeat it
several times. The assessment of lifetime exposure from the gen-
eral environment, or the work place, to a toxic element therefore
requires a different approach. Because toxic elements like Mn are
well retained by bone tissue and may reside there for years to
decades, a non-invasive, non-destructive and without-pain neu-
tron activation based diagnostic technique is being developed and
used to assess Mn. Taking into account the large accumulation of
Mn in bone, an X-ray fluorescence spectroscopy or neutron-based
spectroscopy methods are being considered as novel non-invasive
tool for assessing Mn exposure and toxicity [13]. However, low
energy of Mn X-rays, together with their relatively low fluorescence
yield make neutron activation analysis (NAA) the more promis-
ing approach. This non-invasive in vivo technique can (i) measure
Mn amounts in humans, (ii) contribute to limited human data,
regarding current knowledge of Mn metabolism and (iii) possi-
bly be helpful in early detection of accumulation resulting from
over-exposure in industrial and environmentally exposed popula-
tions. Of many different techniques applied e.g. atomic absorption
spectrometry and inductively coupled plasma atomic emission
spectrometry or mass spectrometry to the determination of total
Mn content, NAA is a prime example of a non-invasive analytical
technique. This method is extremely valuable because of its high
specificity and sensitivity for very low concentrations of Mn, as well
as several other elements. In contrast to other commonly used tech-
niques, in the case of NAA the sample is not permanently damaged
and can be saved for further analysis. The technique is also useful
to benchmark the accuracy of results obtained by other analytical
methods.
The 55 Mn(n,)56 Mn reaction is used for the in vivo neutron acti-
vation analysis (IVNAA) technique. 55 Mn, a stable isotope (natural
abundance is 100%) undergoes thermal neutron absorption with a
high capture cross section of 13.3 barn, producing a radioactive iso-
tope 56 Mn, which decays with a half-life (T1/2 ) of 2.58 h, emitting
a ␥-ray of energy 0.846 MeV having intensity (I␥ ) of 98%. The sub-
stantial amount of magnesium present in the human body produces
27 Mg through the 26 Mg(n,)27 Mg reaction, which is expected to
cause spectral interference, since 27 Mg emits a ␥-ray of 0.843 MeV
with I␥ = 72%, close in energy to the ␥-ray of 56 Mn. However, the
difference in half-lives of the two is advantageous in differentiat-
ing their contributions as 27 Mg is comparatively short lived with a
T1/2 of 9.46 min. The 48 Ca(n,)49 Ca reaction is also studied for nor-
malization and quantification, 49 Ca decays with a T1/2 of 8.7 min
and a ␥-ray of energy 3.08 MeV with I␥ = 91% [14].
IVNAA of Mn performed using a Tandetron accelerator is based
on the neutron production reaction, 7 Li(p,n)7 Be. The Q value of the
endothermic nuclear reaction is 1.64 MeV and the threshold energy
is 1.88 MeV. The possible maximum energies of neutrons from
protons of 2 MeV and 2.3 MeV are 0.23 MeV and 0.57 MeV respec-
tively along the beam axis i.e. 0◦ and the total neutron yield from
2.3 MeV protons is 5.25 times that 2.0 MeV protons [15]. There-
fore, the maximum energy of neutrons is such that the potentially
interfering reactions with iron 56 Fe(n,p)56 Mn (Eth = 2.97 MeV) and
57 Fe(n,d)56 Mn (E = 8.48 MeV) [14], cannot take place.
th
The hand bone is chosen for in vivo neutron activation
because bone tissue has the highest concentration of Mn with
205
approximately 50% of the total body Mn content [3]. In addition,
from number of animal studies it is evident that bone is a long
term storage site for manganese and to measure a hand bone is
technically convenient compared to complexity of using neutron
activation method at other sites of body. Also, the effective radiation
dose to the subject is much lower when only the hand is exposed to
the neutron beam as opposed to a whole body irradiation. Earlier
studies done at our laboratory demonstrated the accumulation of
Mn in human hand following the inhalation of the element in occu-
pational exposures, and that it is measurable using IVNAA [16–20].
We have conducted a few feasibility tests and measurements on
occupational subjects but the levels of Mn in the control subjects
were below the MDL of this previous system. After the feasibility
studies confirmed the measurement of Mn, the main effort was to
improve the MDL values further so as to permit measurements of
(i) subjects with low exposure and (ii) non-exposed individuals as
well. The aim of this study was to explore the factors required to
improve the MDL for the in vivo set-up at McMaster University.
The main variables of accelerator-based IVNAA are the energy of
the projectile and it’s current, the irradiation time, counting strat-
egy and concentration of elements in tissue equivalent hand bone
phantoms, irradiation cavity and counting system. For this study, all
the possible parameters were explored and their effects were stud-
ied. Also, this study was done using on an upgraded (in-house) pulse
processing system [21]. This article presents the resulting improved
calibration curves and MDL with a comparison to previous stud-
ies aimed at making the technique useful for low or non-exposed
subjects.
Materials and methods
Tissue equivalent hand-bone phantoms containing a solution
of Mg, Ca, Na, Cl and Mn in dil. HNO3 were prepared in 250 ml
low density polyethylene (LDPE) Nalgene bottles. The mass of Mn
in the phantoms varied from 0.0 to 12.4 mg while the masses of
Ca, Na and Cl were kept the same as the phantoms used in earlier
measurements [19], based on the data available in ICRP 23 [22].
The mass of Mg was 500 mg in each phantom based on estimations
made from recent Mg measurements conducted as part of a study
in our Alzheimer’s disease patients [23]. Table 1 lists the details of
the element masses used with their respective reaction, half-life
and their thermal neutron capture cross-section.
As mentioned earlier for this in vivo study, an accelerator-based
neutron source was used followed by delayed neutron activation
analysis. The hand bone phantoms were irradiated in the accel-
erator beam following one of two protocols of different proton
energy, current and irradiation time, i.e. (i) 2.0 MeV, 300 ␮A, 3 min
and (ii) 2.3 MeV, 400 ␮A, 45 s. To measure the gamma activity in the
phantoms, a system of eight 10.2 cm × 10.2 cm × 40.6 cm and one
10.2 cm × 10.2 cm × 10.2 cm NaI detectors, arranged in an array of
4␲ geometry, was used [24]. The irradiated phantoms were then
placed inside the gamma counting cavity and the ␥-ray spectra
from the activated phantoms were acquired. The counting strategy
for each phantom was chosen in such a way as to get the contri-
bution of both short- and long-lived isotopes of interest. Average
cooling time between the end of irradiation and the start of the
count was about 30 s. Counting was done in variable time cycles;
for the first 1.5 h the counting was done in 10 min cycles to account
for Ca (T1/2 = 8.7 min) and Mg (T1/2 = 9.46 min) and after 1.5 h the
counting was continued with 30 min cycles for 5–7 h. A typical
gamma spectrum acquired for the 12.4 mg of Mn phantom at three
different time intervals after the irradiation is shown in Fig. 1,
where the strong gamma peaks with their respective elements are
also labeled. Detailed neutron dosimetry of the irradiation cavity
was also carried out in a separate study using a tissue-equivalent
206 C. Bhatia et al. / Journal of Trace Elements in Medicine and Biology 31 (2015) 204–208

Table 1
The masses of elements used in phantom solution with their respective reaction, half-life and cross section.

Element Reaction Mass (g) (ICRP) Compound used Half-life Capture cross section (mb) [14]

Manganese 55
Mn(n,)56 Mn Variable Mn standarda 2.58 h 13,360 ± 50
Magnesium 26
Mg(n,) 27 Mg 0.5c Mg standardb 9.46 min 38.2 ± 0.8
Sodium 23
Na(n,) 24 Na 1.25 Na, NO3 14.997 h 530 ± 5
Chlorine 37
Cl(n,)38 Cl 1.19 NH4 Cl 37.24 min 433 ± 6
Calcium 48
Ca(n,) 49 Ca 14.9 Ca(NO3 )2 8.7 min 1090 ± 140
a
996 ± 4 mg L−1 .
b
1000 ± 2 mg L−1 .
c
Calculated.

proportional counter [25]. The signals from each of the nine detec-
tors were recorded through digital multi-channel analyzer and
gamma peak-fitting was done to estimate the respective net peak
areas of Mn and Ca, corrected for transfer and decay time. The slope
of the calibration curve was obtained from the ␥-ray peak analysis
of induced activity of 56 Mn in the irradiated phantoms i.e. a cali-
bration standard is required to quantify the in vivo ␥-ray signal and
thus its amount in the hand bones of a subject under study.
Results and discussion
observed MDL is an improvement on the previous measurement by
a factor of about 2.3.
As compared to the previous measurements carried out dur-
ing the last study done by Aslam et al. at McMaster, the major
changes adopted in the present study are: (i) Phantoms: The mass
of Mg is doubled (500 mg) based on the estimations made from
Mg measurement in the Alzheimer’s disease study [23]. (ii) Of
the previous measurements done, Mg interference was taken into
account only in Ref. [19]. (iii) Counting strategy: Long counting in
cycles with time varying from 10 to 30 min helps us in deciding

The calibration curves obtained for the two different protocols

studied are shown in Fig. 2(a) and (b). Linear regression curves were
fitted for both data sets and slopes of the curves were used to deter-
mine the net count rate and hence the mass of Mn. After applying
the required corrections for the transfer and irradiation time, the
count rate of Mn was normalized to Ca in order to make the accuracy
of the measurement independent of factors such as target quality,
variation in beam current or positioning of phantom in the cav-
ity, and thickness of overlying soft tissue i.e. expressing Mn mass
per unit bone mass. The Mn net count sensitivity for 2.3 MeV and
2.0 MeV protocols along with the detection limits and a comparison
with available literature data are presented in Table 2. The obtained
sensitivity was 3.2 times the sensitivity of the earlier measure-
ments done at McMaster University [19]. The minimum detection
limit (MDL) of Mn content for our set-up, was estimated as twice
the uncertainty in the peak area of the zero mg Mn phantom with
added mixture, after conversion to mass using the Mn sensitivity
obtained from the respective calibration curves. The revised MDLs
obtained in this study, along with a comparison with previous val-
ues obtained, are presented in Table 2. As shown in this table, the

Fig. 2. (a) Linear calibration curve obtained from the measurements of irradiation
Fig. 1. Typical gamma spectrum for 12.4 mg of Mn along with Mg, Ca, Na, Cl after of Mn mixed hand phantoms for Ep = 2.0 MeV, 300 ␮A, 3 min and (b) for Ep = 2.3 MeV,
irradiation for three different time intervals. Elements with their respective energy 400 ␮A, 45 s. The slope of the curve converted to mass units is used to determine
of gamma lines are shown. the Mn mass in human hands.
 C. Bhatia et al. / Journal of Trace Elements in Medicine and Biology 31 (2015) 204–208 207

Table 2
Summary of observed detection limits for in vivo neutron activation analysis of Mn and comparison with available literature data.

Authors Description of Irradiation Irradiation/cooling/ Local equivalent Detector Sensitivity MDL (␮gMn/gCa)
facility parameters counting time dose (mSv) H (counts/mg)

Zhang et al. [27] Brookhaven Reactor power 4 min/4 min/5 min N/A Pair of Ge(Li) 3827 ∼0.5 ␮g Mn
Medical 2 MW
Research
Reactor
Arnold et al. [16] McMaster Ep = 2.00 MeV; 10 min/6 min/60 min 109 Pair of NaI(Tl) N/A 4.6
KNVan de Graff Ip = 42 ␮A
Byun et al. [18] McMaster Ep = 2.00 MeV; 3 min/30 s/30 min 20 4␲ NaI(Tl) 12,046 5.8
KNVan de Graff Ip = 50 ␮A
Pejovic et al. [17] McMaster Ep = 2.00 MeV; 3 min/30 s/30 min 11 4␲ NaI(Tl) 8879 7.3
KNVan de Graff Ip = 50 ␮A
Aslam et al. [19] McMaster Ep = 2.00 MeV; 3 min/65 min/30 min 18 4␲ NaI(Tl) 17,586 1.6
Tandetron Ip = 100 ␮A
This work (2013) McMaster Ep = 2.00 MeV; 3 min/30 s/30 min 59.7 4␲ NaI(Tl) 56,314 ± 1674 0.71 (1.17%)
Tandetron Ip = 300 ␮A
This work (2013) McMaster Ep = 2.30 MeV; 45 s/30 s/30 min 450.5 4␲ NaI(Tl) 230,459 ± 3101 0.67 (1.01%)
Tandetron Ip = 400 ␮A

a statistically optimal counting interval allowing for decay from Conclusions

short-lived interfering elements, as well as restricting the duration
of the counting interval so as to be acceptable to human sub-
jects. The MDLs obtained from the counting done utilizing short
time intervals over the hours is shown in Fig. 3; the MDL minima
obtained in the region of 50–80 min are used to calculate an inverse
variance weighted mean (IVWM) of the MDLs of the two proto-
cols. (iv) Data acquisition system (DAQ): The acquisition system
was upgraded since the time of the last measurements and details
were discussed previously [21]. (v) Accelerator current: The proton
currents for the 2.0 MeV and 2.3 MeV protocols used are 300 ␮A and
400 ␮A respectively, 54 mC at 2.0 MeV gives an equivalent dose of
59.5 mSv, whereas 18 mC at 2.3 MeV gives an equivalent dose of
450.5 mSv [25].
Since, in the present measurements Mg content is doubled,
which increases uncertainty, the sensitivity does not increase pro-
portionally to the increase in current i.e. a factor of three. Of the two
protocols investigated, the combination of: 2.0 MeV, 300 ␮A, 3 min
protocol is recommended for further measurements of in vivo Mn
measurements studies as compared to the 2.3 MeV, 400 ␮A, 45 s
protocol. With the former, we saw a much lower dose-equivalent
and a relatively minor difference in MDL. The factors that could be
considered in improving the detection limit further are the dose-
rate to the subjects; should we manage to do longer counting of
(say) 1–2 h, we would expect the MDL to improve further by a factor
2–3 as can be seen from Fig. 3.
This work presents the study of key components resulting in
improvement of the MDL for an accelerator based non-invasive,
in vivo technique used for the measurement of Mn in tissue equiv-
alent human hand-bone phantoms. The technique could serve
as a quantitative and analytical approach for measuring elevated
amounts of Mn exposures in humans.
The important factors studied which have been taken into
account for the estimation of MDL for Mn in bone include opti-
mally selected beam parameters for irradiation, using an upgraded
DAQ system, a revised more realistic mass of magnesium, use of
higher neutron fluence by increasing the current, and estimation
of dose equivalents and counting strategy. Combining these fac-
tors has reduced the overall detection limit by a factor of nearly
2.3 with our present experimental set-up at McMaster University.
A comparison of MDL with previous measurements has also been
presented. The present MDL values are close to the reference value
of 0.63 ± 0.30 ␮gMn/gCa for the human skeleton [26]. Work is in
progress to extend this study for low (<2 mg) mass of Mn phantoms.
Thus, the improved MDLs revealed in this work could be of pro-
found importance for use in the measurement of occupational and
environmental exposures to Mn, and hence provide a potentially
improved assessment of health risks, in clinical studies aimed at
better understanding the mechanism of Mn toxicity.
Conflict of interest

The authors declare that there are no conflicts of interest.

Acknowledgements

The authors extend their thanks to Justin Bennett, accelera-

tor operator and Scott McMaster, accelerator manager, McMaster
University for their assistance in the operation of the Tandetron
accelerator and S.D. Molla for providing us the recent dose equiva-
lent data. The project was funded by a research grant provided by
AECL, Canada.

References

Fig. 3. Minium detection limit (␮gMn/gCa) spectrum observed as a function of time
for the 2.0 MeV proton energy, 300 ␮A current and irradiated for 3 min.
[1] Cotzias GC. Manganese in health and disease. Physiol Rev 1958;38:503–32.
[2] Food Nutrition Board, Institute of Medicine. Dietary reference intakes for
vitamin A, vitamin K, boron, chromium, copper, iodine, iron, manganese,
molybdenum, nickel, silicon, vanadium, and zinc. Washington, DC: National
Academy Press; 2001. p. 394–419.
[3] Pearson GF, Greenway GM. Recent developments in manganese speciation.
Trends Anal Chem 2005;24:803–9.
208 C. Bhatia et al. / Journal of Trace Elements in Medicine and Biology 31 (2015) 204–208
[4] Aschner M, Erikson KM, Dorman DC. Manganese dosimetry: species differences
and implications for neurotoxicity. Crit Rev Toxicol 2005;35:1–32.
[5] Gerber GB, Leonard A, Hantson P. Carcinogenicity, mutagenicity and terato-
genicity of manganese compounds. Crit Rev Oncol Hematol 2002;42:25–34.
[6] Schroeder WH, Dobson M, Kane DM. Toxic trace elements associated with air-
borne particulate matter: a review. J Air Pollut Control Assoc 1987;37:1267–85.
[7] Crossgrove J, Zheng GW. Manganese toxicity upon overexposure. NMR Biomed
2004;17:544–53.
[8] Clewell HJ, Lawrence GA, Calne DB, Crump KS. Determination of an occupational
exposure guideline for manganese using the benchmark method. Risk Anal
2003;23:1031–46.
[9] Bowler RM, Koller W, Schulz PE. Parkinsonism due to manganism in
a welder: neurological and neuropsychological sequelae. Neurotoxicology
2006;27:327–32.
[10] Andersen ME, Gearhart III JM, Clewell HJ. Pharmacokinetic data needs to sup-
port risk assessments for inhaled and ingested manganese. Neurotoxicology
1999;20:161.
[11] Roels HA, Ghyselen P, Buchet JP, Ceulemans E, Lauwerys R. Assessment of the
permissible exposure level to manganese in workers exposed to manganese
dioxide dust. Br J Ind Med 1992;49:25–34.
[12] Finley BL, Santamaria AB. Current evidence and research needs regarding the
risk of manganese-induced neurological effects in welders. Neurotoxicology
2005;26:285–9.
[13] Zheng W, Sherleen XF, Dydak U, Cowan DM. Biomarkers of manganese intoxi-
cation. Neurotoxicology 2011;32:1–8.
[14] Tuli JK. The Nuclear Wallet Cards. 8th ed. USA: NNDC, BNL; 2011.
[15] Lee CL, Zhou XL. Thick target neutron yields for the 7 Li(p,n)7 Be reaction near
threshold. Nucl Instrum Methods Phys Res B 1999;152:1–11.
[16] Arnold ML, McNeill FE, Stronach IM, Pejović-Milić A, Chettle DR, Waker A. An
accelerator based system for in vivo neutron activation analysis measurements
of manganese in human hand bones. Med Phys 2002;29:2718–24.
[17] Pejović-Milić A, Byun SH, Chettle DR, McNeill FE, Prestwich WV. Development
of an irradiation facility for in vivo neutron activation analysis of manganese in
human bone. J Radioanal Nucl Chem 2006;269:417–20.
[18] Byun SH, Pejović-Milić A, Prestwich WV, Chettle DR, McNeill FE. Improvement
of in vivo neutron activation analysis of Mn using a 4␲ NaI(Tl) detector array. J
Radioanal Nucl Chem 2006;269:615–8.
[19] Aslam, Pejovic-Milic A, Chettle DR, McNeill FE. Quantification of manganese in
human hand bones: a feasibility study. Phys Med Biol 2008;53:4081–92.
[20] Pejović-Milić A, Aslam, Chettle DR, Oudyk J, Pysklywec M, Haines T. Bone man-
ganese as a biomarker of manganese exposure: a feasibility study. Am J Ind
Med 2009;52:742–50.
[21] Matysiak W, Atanackovic J, Katalmohseni H, Byun SH, Inskip MJ, Prestwich WV,
et al. In-vivo neutron activation analysis for aluminum in bone: system upgrade
and improved data analysis. AECL Nucl Rev 2013;2:27–32.
[22] ICRP. Report of the Task Group on Reference Man, No. 23. Oxford: Pergamon
Press; 1975.
[23] Chettle DR, Cowan D, Pejović-Milić A, Priest ND, Matysiak W, Atanackovic JZ,
et al. A pilot study measuring aluminium in bone on Alzheimer’s and referent
subjects: work in progress. Winchester: Tenth Keele Meeting on Aluminium;
February 2013.
[24] Byun SH, Prestwich WV, Chin K, McNeill FE, Chettle DR. 4␲ NaI(Tl) detec-
tor array for in vivo neutron activation analysis. IEEE Trans Nucl Sci
2006;53:2944–7.
[25] Byun SH, Pejović-Milić A, McMaster S, Matysiak W, Aslam, Liu Z, et al. Dosi-
metric characterization of the irradiation cavity for accelerator-based in vivo
neutron activation analysis. Phys Med Biol 2007;52:1693–703.
[26] Bush VJ, Moyer TP, Batts KP, Parisi JE. Essential and toxic element concentrations
in fresh and formalin-fixed human autopsy tissues. Clin Chem 1995;41:284–94.
[27] Zhang RQ, Ellis KJ. In vivo measurement of total body magnesium and man-
ganese in rats. Am J Physiol Regul Integr Comp Physiol 1989;257R:1136–40.