Professional Documents
Culture Documents
PII: S0379-0738(16)30020-2
DOI: http://dx.doi.org/doi:10.1016/j.forsciint.2016.02.007
Reference: FSI 8314
Please cite this article as: K. Krishan, P.M. Chatterjee, T. Kanchan, S. Kaur, N. Baryah,
R.K. Singh, A review of sex estimation techniques during examination of skeletal
remains in forensic anthropology casework, Forensic Science International (2016),
http://dx.doi.org/10.1016/j.forsciint.2016.02.007
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
Acknowledgments
Acknowledgements
This review article on methodological issues has been the outcome of the
authors’ experiences (KK and TK) in the field of sex estimation and forensic
anthropology of over a decade. It is hoped that the results of these experiences would
t
ip
be fruitful to future researchers. The corresponding author wishes to acknowledge
cr
PURSE Grant), Panjab University, Chandigarh, India for providing platform for writing
us
such exhaustive work in forensic anthropology.
an
M
d
p te
ce
Ac
Page 1 of 39
*Title Page (with authors and addresses)
Review article
t
ip
Authors with Affiliations
cr
Preetika M. Chatterjee, 1, MSc (UGC Junior Research Fellow)
us
Tanuj Kanchan2, DFM, MD (Associate Professor)
1
Department of Anthropology, Panjab University, Chandigarh, India
2
Department of Forensic Medicine and Toxicology, Kasturba Medical College (Affiliated
p
3
Department of Forensic Medicine, Pt. J.N.M. Medical College, Raipur, Government of
Chhattisgarh, India
Ac
Funding: Centre for Advanced Study (CAS) in Anthropology, DST PURSE grant,
Department of Anthropology, Panjab University, Chandigarh, India
Conflict of interest- The authors declare that there is no conflict of interest regarding
this manuscript
Page 2 of 39
Corresponding Author:
t
+91-9876048205 (Mobile)
ip
cr
us
an
M
d
p te
ce
Ac
Page 3 of 39
*Highlights (for review)
Highlights
casework
radiographic methods
t
ip
Direct methods of sex estimation are found to be more reliable than the other
cr
methods
GM and DSP methods are emerging as valid and widely used techniques in
us
forensic anthropology
an
M
ed
pt
ce
Ac
Page 4 of 39
*Manuscript (without author details)
Review article
ABSTRACT
t
anthropology casework, and requires foremost consideration in the examination of
ip
skeletal remains. Forensic anthropologists frequently employ morphologic and metric
cr
methods for sex estimation of human remains. These methods are still very
us
imperative in identification process in spite of the advent and accomplishment of
population data. These methods however, are less reliable owing to high variance
M
and indistinct landmark details. The present review discusses the reliability and
the bones and hence, such direct methods of sex estimation are considered to be
ce
more reliable than the other methods. Geometric morphometric (GM) method and
Diagnose sexuelle probabiliste (DSP) method are emerging as valid methods and widely
Ac
Besides, the newer 3D methods are shown to exhibit specific sexual dimorphism
better methodologies for sex estimation as well as re-evaluation of the existing ones
will continue in the endeavour of forensic researchers for more accurate results.
Page 5 of 39
Keywords
1. Introduction
t
Forensic anthropology deals with the identification of human remains in a
ip
legal context. It refers to the application of the science of physical/ biological
cr
anthropology in forensic context [1,2]. Forensic anthropology thus, is a
us
multidisciplinary field that deals with a wide spectrum of issues ranging from
(EAFS) and other associations and societies of forensic sciences and legal
ce
Page 6 of 39
The need for identification of remains arises in both natural as well as man-
made disasters where the dead bodies may be highly decomposed, dismembered
or skeletal remains is a humanitarian need and imperative for legal recourse [4].
t
bodies and human remains are brought by the criminal investigating agencies for
ip
examination. Bones are the framework of the vertebrate body, and thus the source of
cr
information about the adaptive mechanisms to its environment [5]. The study of
us
bones and their anatomical measurements not only gives information about their
origin but also ancestry, sex, stature and age at death thus, helping in establishing
an
the identity of an individual. Forensic anthropologists are thus, involved in death
investigations, and assist in determining the identity, cause and manner of death.
M
Identification in forensic anthropology is a two-step procedure that includes
assessment of group and individualistic traits. Initial estimation of group traits such
ed
as age, sex, stature, and ancestry serves as an important clue to identification that is
person such as moles, scars, tattoo marks, or any abnormality and deformity evident
ce
narrowing down the search from a group to an individual. The comparison process
especially in the mass fatality incidents is quite difficult to achieve due to the lack of
is vital in estimation of age, ancestry and stature as there are observable differences
3
Page 7 of 39
in ageing and growth patterns between sexes, and variations in morphological traits
unknown adult skeleton than a neonate or a juvenile one owing to the insignificant
sex differences exhibited in the skeletal morphology during the pre-pubertal period.
Geometric and statistical shape analysis methods have been utilized by some
t
studies for sexing during pre-pubertal period, and are largely focussed on pelvis and
ip
skull [10]. All in all it is nearly impossible to sex the sub-adult skeleton on the basis of
cr
the morphological features. On the other hand, all parts of the adult skeleton varying
us
from small bones of hand and foot [11], long bones [12], patella [13], pelvis [14,15],
sternum [16] and cranium and its parts [17-19] have been utilized for sex estimation
an
in different populations worldwide. The reliability and accuracy of sex prediction
depends directly on the anatomical region of the available skeletal remains [20].
M
2. Sex estimation – An overview of factors influencing skeletal morphology
Biological sex represents the genotype while gender is the phenotype of the
ed
based on reproductive role and attributes [21], that is shown to exhibit inter and intra
pt
dissimilarities among sexes which ultimately are depicted in body tissues [22]. Thus,
the ultimate function adopted by the tissue rules the final appearance and various
Ac
responses of the same. Biomechanical loading and stress along with the functions
undertaken by the body part determines the morphology of that part [1].
Besides, secular changes are observed in the size and height [23], and in the
size of individual bones [23-25] between generations. Males exhibit larger stature,
more robust cranial and facial features, greater muscularity, strength and speed as
compared to females [26]. Differences in pre- and post-natal hormone levels, growth
4
Page 8 of 39
rates and diseases too have differential effects in both sexes [27]. Age related
changes are observed in skeletal material that at times can cause erroneous
masculine cranial features while young men show more gracile and feminine cranial
features [28]. While sexing the human remains especially the skulls, the investigator
t
should also take into consideration the pedomorphism in the human skeleton. The
ip
human adult skulls may retain some juvenile traits which may result in erroneous
cr
conclusions. The female skulls are much more pedomorphic than the males. After
us
the onset of puberty, the differences between the male and female skull become
clearer as the male skull develops some adult characteristics, however, the female
Sexing techniques are primarily directed to the pelvis for exhibiting the reproductive
ed
differences in the bone, and the cranium in which the variability in size and
morphologies among the sexes is represented at its best [31]. Apart from the pelvis
pt
and skull, long bones are the most actively studied skeletal elements [32,33]
ce
followed by ribs [34,35], vertebrae [36], clavicle [37], sternum [38,39] and the bones
in the body for its variable appearance and function in males and females. This
explains the greater attention received by some bones in sex prediction over others
Page 9 of 39
Two principle approaches in sex estimation are based on the methods of
Iscan and Derrick [43] and Phenice [44]. These approaches form the basis of
in the existing methods of sex estimation and introduction of new methods from time
to time, there is still a great demand for generating population specific standards
t
accurately using more reliable methods [31,45]. Sex estimation methods have been
ip
developed using almost every part and element of the skeleton with varying degree
cr
of accuracies.
us
The methods used commonly for sex estimation can be broadly classified as
techniques like DNA fingerprinting [4]. Molecular methods are highly sophisticated,
M
and demand higher level of skill and latest equipments. These methods have greater
degree of reliability, but are complicated, invasive, highly expensive, and time
ed
consuming [46,47].
acceptable in forensic setting which can vary depending upon the condition of
Ac
remains available for examination [1]. Burning of corpse causes shrinkage of bones
allocate morphologic features that are sex specific [46]. Some studies [48-51]
showed that although, the heat related fragmentation, warping and the dimensional
changes in the bones affect the sex determination process, nevertheless, high
accuracy can be achieved with new osteometric references for sex determination in
6
Page 10 of 39
human remains. However, in cases of badly decomposed or incomplete specimens
which cannot provide the required information sufficient enough to allow diagnosis,
simpler methods are usually applied initially, and to adopt a higher method, the
t
skeleton, degree of clarity of features present and precision required in a particular
ip
case. Chronology and geographic origin forms the basis of selecting a parameter to
cr
be included in any sex prediction model [52]. The sexual probability of an individual
us
and the bone parameters are related to one another in a non–linear manner [53,54].
however, are greatly influenced by the level of subjectivity [56,57]. Many features of
ed
sex assessment like glabellar prominence, shape of the pubic bone or mandibular
ramus flexure can only be assessed morphologically [58,59]. These features are
pt
difficult to measure, show inter and intra observer errors, classification difficulties and
ce
bones or incomplete skeletons. Walker [54] used the criteria given by Buikstra and
Ubelaker [45] for visual scoring on the rating scale of 1-5 and reported a sexing
accuracy of 90% for skulls. The features were also subjected to univariate analysis
with a sexing accuracy of 70%, and multivariate analysis with a sexing accuracy of
80%-88%. In the study conducted by Loth and Henneberg [60], the accuracy
achieved was 81% using morphologic analysis of symphyseal region and anterior
7
Page 11 of 39
body of the mandible in juveniles, however, this accuracy decreased to 64% in a
blind test conducted by Scheuer [61] using the same method, exhibiting great
dissimilarity among the two studies. These observations emphasise on the fact that
t
Metric studies are based on the basic principle of variability in male and
ip
female dimensions and mostly utilize different statistical methods to derive models/
cr
equations that can be used for sexing of individuals. Numerical results obtained from
us
metric studies are easier to assess and interpret [47,63-67]. In a study by Franklin et
al. [68] on five tribes of South Africa depicting inter-population differences; nine
an
measurements on the mandible presented male and female differences among
logistic regression analysis and discriminant function analysis for sexing [69,70]. The
accuracy in sexing obviously may vary based on the statistical method utilized. Apart
pt
from sexing based on measurements directly, indices have been derived for their
ce
such as body build and stature have not yielded large sex differences [16,71-72].
Ac
other classical osteometric studies [73,74]. While the accuracy ranges from 85-95%
the tunes of 96% has been reported using virtual morphognostic sexing methods
8
Page 12 of 39
[85]. Digital radiographs provide an alternative and accurate measuring technique
decomposed and degraded samples where maceration cannot be tried prior to the
analysis. Radiographs taken directly on bones [86] or on living individuals [85] have
t
obtained from minimum supero-inferior neck diameter and maximum head diameter
ip
of femur. Congruent inferences have been reported in black Malawians by Igbigbi
cr
and Msamati [87] from digital radiographs and X-ray films in a retrospective study of
us
496 pelvis with in the age range of 18-70 years where femoral head diameter was
used to determine sex. Both the diameters i.e. vertical and antero-posterior
an
diameters were considered and identification and demarking points used for
Imaging (MRI)
pt
CT and MRI are non-invasive methods having vital role in actual forensic
ce
anthropological methods making them more responsive and accurate in the ever
Ac
greater sciatic notch to construct a three dimensional image for analyzing sex
Page 13 of 39
reported that sexing could be done for deeper and more resistant parts using this
population applied an advanced method on the data derived from 3-D computed
tomography. This method studied the 3-D anatomical land marks in a digital format
t
instead of any physical measurements on a sample of 91 Thai dry skulls. The study
ip
revealed that Thai males were larger mainly in the parameters such as maximum
cr
cranial length, basion-bregma height, nasion-basion length, nasion-basion breadth
us
and bizygomatic breadth. In another study, volume–rendered cranial CT scans were
identified as males and 40 of 100 crania as females with an accuracy level of 96%.
M
Arcus superciliaris alone allowed the most accurate sex estimation among males
rates among the two sexes and very low inter-observer biases were observed.
morphology of rigid structures having curves and bulges which cannot be analysed
quantify shape since 1980’s have gained popularity in physical anthropology in the
morphological characters can be avoided at par with its ability to include surface
topography using x, y coordinates [56]. The analysis is done in such a manner that
the effect of size as a parameter is eliminated and even subtle differences are
10
Page 14 of 39
considered increasing the accuracy level manifold [83,88,96]. In this method the
sample is photographed and put into a computer [83], the landmark coordinates are
then digitized and analysed using generalised procrustes analysis in which these are
represented as points in the shape space of Kendall [97] through the process of
scaling and alignment. It uses centroid size (CS) in which all the specimens are
t
scaled. Subsequent multivariate analysis is done using superimposed coordinates
ip
and CS. Principal Component Analysis with covariance matrix is done on
cr
generalised procrustes analysis variables. A multivariate analysis of covariance
us
(MANCOVA) is then done using Principal Component Analysis scores to check
significance of size and sex on shape of male and female in each group. These
an
groups are further subjected to Discriminant Analysis of covariance if satisfactory
16 craniofacial landmarks near mandible and alveolar bone on 118 adults from W.M.
ed
Bass donated collection and Forensic Data Bank. The study observed a significant
influence of sex on the shape but not on size in American Black and Whites thus,
pt
contrasting the findings of Rosas and Bastir [95] who reported great influence of size
ce
and sex on shape in craniofacial region using 2D geometric morphometrics for sex
methods. A study by Styen et al. [56] using TPSDIG program to digitize landmarks
observed narrower greater sciatic notch shape in South African black males that was
otherwise wider in black and white females with scattered shape variation in white
males rendering it unfit to be used for sex estimation. With Geometric Morphometric
Page 15 of 39
reported with both semi-landmark and multivariate statistical methods for the greater
sciatic notch while the accuracy with shape variables for ischio-pubic complex was
93.4% and 90.1% for DFA. Although landmark based Geometric Morphometric
Analysis appears to be the most appropriate method, its biggest disadvantage is that
the available number of landmarks may not be sufficient to define the shape of a
t
specimen. Besides, it remains a time consuming procedure [83].
ip
4.2.4 Diagnose Sexuelle Probabiliste method (DSP)
cr
Murail et al. [99] devised a virtual method of sex determination based on
us
metric database derived from 2040 hipbones of known sex from twelve different
reference populations [100]. The database is based on the hypothesis that all the
an
modern human populations share a common pattern of sexual dimorphism in their
hip bone. In order to validate this method, four different geographical areas namely
M
Europe, Africa, Asia, and North America were included along with one to three
individuals with known age and sex were taken into account. In this method, the
Variables were selected on the basis of their discriminant power and preservation
ce
rate. This pooled sample from worldwide collections was utilized to develop DSP.
Sex diagnosis was based on any combination of at least four variables among those
Ac
which have been proposed. Sex determination was done by comparing the
males and females. Chapman et al. [100] examined the accuracy of DSP as
compared to other methods of virtual sex determination and its accuracy for both
hipbone and pelvic girdle to determine its utility in forensic casework. They achieved
100% accuracy between the manual and virtual DSP method. DSP was reported to
12
Page 16 of 39
be a robust method and applicable in forensic casework including both hipbone and
pelvic girdle with 100% accuracy. Mestekova et al. [101] tested the reliability of DSP
method on CT images of the hip bone from 52 and 54 females of a modern French
achieved 92.3% accuracy in males and 97.2% accuracy in females. The study
t
supported the view that DSP is a reliable method of sexing based on CT images of
ip
hip bone in French sample too. A large number of combinations make sex
cr
determination of both well preserved as well as damaged sample, a simple and easy
us
criteria, thus, making DSP one of the methods of choice for forensic purposes in this
regard.
an
5. Analysis and models employed in sexing of remains
regression analysis and discriminant function analysis for sexing. The accuracy in
ed
sexing obviously may vary based on the statistical method utilized. Of all the
statistical methods, discriminant function analysis (DFA) remains the most widely
pt
subjective criteria for sex estimation and is simple to use without any prior
discriminant functions have a vital role in estimation of sex [31]. DFA helps in
assigning any subject having unknown identity to one of the two or more groups
Page 17 of 39
are highly population specific [11,63,102,103-105] and the percentage accuracy vary
from one measurement set to another [31]. It is applied to analyse a set of axes
which give highest possibility of discrimination among two or more groups [106].
Firstly, it distinguishes among the groups and computes classification scores. After
that the individuals are classified according to the group for which they show highest
t
classification score. These are then cross validated to evaluate the accuracy
ip
achieved. In cross validation analysis, the cases are classified with the functions
cr
derived from all the cases other than the case that is to be classified. Thus, a series
us
of analysis is performed; excluding one individual at a time i.e. leave one out cross
validation [96].
an
This method was used in the studies carried out by Thieme and Schull [107],
Hanihara [108], and Giles [109]. Hanihara [108] reported 90% accuracy with DFA on
M
a Japanese sample using just three cranial dimensions. Degree of resistance to
forensic settings. Thus, DFA is also utilized on robust skeletal elements as femur
pt
[63,73,105], calcaneous and talus [112]. The results of DFA are highly dependent on
ce
sample size, robusticity and sexual dimorphism of the population for which the
discriminant functions have been developed [54]. Sakaue [113] reported distal
Ac
stepwise DFA while a study by Mall et al. [114] reports humeral head to be a reliable
are attributed to population differences in sex estimation. Bidmos and Dayal [115]
used univariate and multivariate DFA for sex estimation using sixty White South
African tali and reported accuracy of 85% with length measurements giving more
14
Page 18 of 39
reliable sexual dimorphism. DFA has been used to determine sex from bones such
cephalometry [87].
5.2.2 Fordisc
t
Fordisc utilizes the global database of known sex samples available in
ip
forensic data bank [fdb] to compute discriminant functions for sexing, stature
cr
estimation and race determination using various anthropometric measurements
us
[119,120]. Along with that this software also performs the similar tests using Howells
data on world wide variation among past populations [121,122]. It also provides
an
accurate statistical data best suited for forensic anthropologists. The basis of this
software is the very nature of discriminant function analysis which implies the
M
knowledge of ancestry of an unknown specimen in order to get adequately reliable
results. Fordisc 2.0, an earlier version was able to analyse sex and ancestry as
ed
separate units. With the newer Fordisc 3.0 versions, selection of appropriate groups,
use of large combination of measurements at ones disposal, and use of the whole
pt
forensic data bank added up to the accuracy of the software but was unable to
ce
enhance the reliability. The basis of published discriminant functions for sex
estimation is pp>0.5, and Fordisc 3.0 uses this by default in the study conducted by
Ac
comparability of unknown sample with reference data bank and availability of certain
better results as compared to Fordisc 3.0 software, a few others do not show very
encouraging results. A study shows that the French samples give better results with
Page 19 of 39
individual based discriminant functions [120]. A degree of accuracy of 63.7% on the
whole and 77.8% in French subsample obtained in this study however, does not fulfil
the reliability criteria needed in forensic anthropology case work. One of the major
problems in the study [120] is the non-homogeneous nature of forensic data bank
and group selection. Another criticism raised on this software is related to ancestry
t
assessment [123-125], however, continuous reviews are being done starting from its
ip
basics to its applicability [126-129] in forensic field.
cr
5.2.3 Neural network method
us
Neural network method, also known as multilayer preceptor method, is one of
the least preferred methods in sexing since very high mathematical skills are
an
required to develop models based on this method. This despite of the fact that the
derived models are more efficient in depicting relationships among variables than
M
DFA and other multivariate non-linear methods of sex estimation having greater
prudent model designs [52,53,130]. Instead of an equation this method uses a matrix
ed
having values represented as nodes similar to that of the neural network of human
brain to show the relationship among variables. This network comprises of an input
pt
layer and an output layer having one or several inter-linked layers. All types of
ce
nonlinear functions between input and output can be modelled without any
conducted by Jardin et al. [52], neural network method having one hidden layer
showed higher accuracy of 93.4% with lesser asymmetry among male and females
than compared to other two methods i.e. DFA and linear regression method using
four parameters obtained from upper part of femur. In an earlier study conducted on
patella using the same method, Mahfouz et al. [53] had reported higher sex
16
Page 20 of 39
estimation accuracy of 96% when compared to DFA that reported an accuracy of
90.3%.
6. Molecular methods
Poor preservation and loss of skeletal elements sometimes does not allow
t
molecular methods come to the rescue of the investigators [131]. Molecular sex
ip
typing using PCR multiplex [Genderplex] [132] uses two different sized amelogenin
cr
targets, one STY target and four X chromosomal STRs featuring short length
us
amplicons. It is a very useful technique in sex estimation due to the reason that it
utilizes parallel testing at multiple sites of genomes. These methods are useful in
an
less degraded and high quality DNA sample, however, short amplicons may prove to
depends on the element present for analysis and its preservation state when sexing
unknown skeletal remains and also varies with one measurement set to another.
Ac
Morphological observations are simpler to make but harder to judge as most of them
in the hands of an experienced observer, this method can allow accurate and precise
often suffer from inter and intra-observer discrepancies if landmarks are not properly
defined. To defend ones findings in the court of law, nowadays forensic scientists
17
Page 21 of 39
preferentially use a combination of non– metric and metric methods. Although digital
image processing method for sexing using digital photographs and technical drawing
accuracy [136], inter-population variance needs to be taken care of. This method can
t
more economical. CT and MRI are although expensive methods nevertheless are
ip
useful in actual forensic scenarios, systematic re-evaluation and improvement of
cr
existing anthropological methods pertaining to secular changes in and among
us
populations in a non-invasive manner. Besides, volume-rendered CCT images are
radiological methods and the direct measurement methods [137]. The problem of
M
excluding the assessment of shape variation between landmarks in linear
allowing the researcher to visually identify exact areas causing variation among
pt
specimens. DSP method of sex determination from hip bones has proved to be a
ce
Discriminant function analysis is the most popular statistical model for sex
Ac
variables thus, giving DFA somewhat rigid nature. Further, DFA using Fordisc 3.0
18
Page 22 of 39
version software can provide reliable results only if the geographic origin of the
sample is available in the databank [140]. This is useful in comparison studies where
the sample can be compared with specific databases. Neural network method can
provide more accurate results than DFA but is more complicated and time
consuming. On the other hand, the molecular methods are highly sophisticated,
t
expensive, skill based and invasive, thus used in cases in which classical
ip
anthropological methods are not applicable at all or the results obtained are
cr
insufficiently reliable.
us
8. Present scenario and future perspectives
where well documented skeletal collections are a rare occurrence and those
M
available are not representative of the current population owing to secular changes,
[141]. Many studies however, support the reliability and reproducibility of data
measurements, however, higher degree of variance has been observed in all these
Ac
sexing. New 3D methods can show specific sexual dimorphism patterns not readily
revealed by the traditional methods. However, that does not undermine the
significance of already standardized and widely used methods for sex estimation in
forensic anthropology. In forensic case scenario, quicker and more efficient methods
19
Page 23 of 39
are and will be employed preferentially to analyze skeletal remains for sex estimation
studies. Development of newer and better methodologies for sex estimation as well
t
References
ip
[1] K.R. Burns, Forensic Anthropology Training Manual, 3 rd Ed, Pearson/Prentice
cr
Hall, Upper Saddle River, N.J., 2012.
us
[2] S. Black, Forensic anthropology—regulation in the United Kingdom, Sci. Justice
[3]
an
C. Cattaneo, Forensic Anthropology: developments of a classical discipline in
[5] T.D. White, P.A. Folkens, The Human Bone Manual, Academic press
forensic context in Kosovo—a case history, J. Forensic Sci. 49 (3) (2004) 464–
468.
20
Page 24 of 39
[9] T. D. Stewart, Essentials of Forensic Anthropology, Springfield, Illinois: C.C.
Thomas, 1979.
Forensic Anthropology: 2000 to 2010, (2011). Boca Raton, FL, CRC Press;
2011.
t
[11] E. Gualdi-Russo, Sex determination from the talus and calcaneus
ip
measurements, Forensic Sci. Int. 171 (2007) 151–156.
cr
[12] Kranioti EF, Apostol MA. Sexual dimorphism of the tibia in contemporary
us
Greeks, Italians, and Spanish: forensic implications. Int. J. Leg. Med.
[13]
an
M.R. Dayal, M.A. Bidmos, Discriminating sex in South African blacks using
57-64.
of modern sternum - an autopsy study from South India, Leg. Med. (Tokyo) 16
(2014) 350-356.
(2005) 1310–1314.
21
Page 25 of 39
[18] Y.P. RaghavendraBabu, T. Kanchan, Y. Attiku, P.N. Dixit, M.S. Kotian,
[19] J. Bruzek and P. Murail, Methodology and reliability of sex determination from
t
complementary sciences from recovery to cause of death. Edited by A. Schmitt,
ip
E. Cunha and J. Pinheiro. Humana Press Inc., Totowa, New Jersey, 2006, 225-
cr
242.
us
[20] L.L. Mai, M.Y. Owl, P. Kersting, The Cambridge Dictionary of Human Biology
44 (2001) 25-53.
M
[22] L.L. Kelpinger, Stature, maturation variation and secular trends in forensic
[23] R.L. Jantz, Cranial change in Americans: 1850 – 1975. J. Forensic Sci. 46
(2003). 517-520.
[26] D.W. Frayer, Sexual Dimorphism, Ann. Rev. Anthropol. 14 (1985) 429-73.
22
Page 26 of 39
[27] W.A. Stini, Growth rates and sexual dimorphism in evolutionary perspective.
t
Remains, 2004,Pp. 23-25.
ip
[29] L. Scheuer, S. Black. Osteology, In T. Thompson, S. Black (Ed) Forensic
cr
Human identification- An Introduction, CRC Press, Taylor and Francis Group,
us
Boca Raton, FL, 2007:199-220.
[32] M. Steyn, M.Y. Iscan, Sex determination from the femur and tibia in South
[33] M.Y. Iscan, S. Loth, C. King, D. Shihai, M. Yoshino, Sexual dimorphism in the
ce
[34] M.Y. Iscan, Osteometric analysis of sexual dimorphism in the sternal end of
sexual dimorphism in the sternal end of the rib in a west African population, J.
23
Page 27 of 39
[36] I. Gama, D. Naveg, E. Cunha, Sex estimation using the second cervical
[37] W.F. McCormick, J.H. Stewart, H. Greene, Sexing of human clavicles using
t
(1991) 175–181.
ip
[38] O. Ekizoglu, E. Hocaoglu, E. Inci, M.G. Bilgili, D.Solmaz, I. Erdil, I.O. Can,
cr
Sex estimation from sternal measurements using multidetector computed
us
tomography, Medicine (Baltimore) 93 (2014) e240.
[43] M.Y. Iscan, K. Derrick, Determination of sex from the sarcoilliac joint: A visual
[44] T.W. Phenice, A newly developed visual methods of sexing os pubis, Am. J.
24
Page 28 of 39
[45] J.E. Buikstra, D.H. Ubelakar, Standards for data collection from human
the forensic diagnosis of sex and age from skeletons, HOMO 58(1) (2007) 75-
t
89.
ip
[47] P.A. Lachenbruch, M. Goldstein, Discriminant Analysis: Perspectives in
cr
Biometry, Published by: Int. Biometric Society, Biometrics 35 (1) (1979) 69-85.
us
[48] D. Gonçalves. The reliability of osteometric techniques for the sex
[49]
an
D. Gonçalves, T.J. Thompson, E. Cunha. Osteometric sex determination of
condition of human remains in forensic contexts, Int. J. Legal Med. 2014 (In
ed
(2013):1134-1140.
between neural network and other metric methods to determine sex from the
upper femur in a modern French population, Forensic Sci. Int. 192 (2009)
127.e1–127.e6.
25
Page 29 of 39
shape models and nonlinear classifiers, Forensic Sci. Int. 173 (2007) 161–
170.
[54] P.L. Walker, Sexing skulls using discriminant function analysis of visually
[55] M.A. Bidmos, V.E. Gibbon, G .Strkalj, Recent advances in sex identification of
t
human skeletal remains in South Africa, S. Afr. J. Sci. 106 (11/12) (2010) 238-
ip
243.
cr
[56] M. Steyn, E. Pretorius, L. Hutten, Geometric morphometric analysis of the
us
greater sciatic notch in South Africans, Homo 54 (2004) 197–206.
99 (1996) 473–485.
ed
first few years of life. Am. J. Phys. Anthropol. 115 (2001) 179–186.
(2002) 297–312.
[62] D.E. Walrath, P. Turner, J. Bruzek, Reliability test of the visual assessment of
[63] M.Y. Iscan, P. Miller-Shaivitz, Determination of sex from the femur in Blacks
Page 30 of 39
[64] M. Steyn, M.Y. Iscan, Sexual dimorphism in the crania and mandibles of
[66] S.A. Asala, Sex determination from the head of the femur of South African
t
whites and blacks, Forensic Sci. Int. 117 (2001) 15–22.
ip
[67] E. Pretorius, M. Steyn, Y. Scholtz, Investigation Into the Usability of
cr
Geometric Morphometric Analysis in Assessment of Sexual Dimorphism, Am.
us
J. Phys. Anthropol. 129 (2006) 64–70.
84.e1- 84.e5.
M
[69] G.R. Dabbs, Is Dwight right? Can maximum height of the scapula be used for
[70] G.R. Dabbs, P.H. Moore-Jansen, A method for estimating sex using metric
[71] S.A. Hunnargi, R.G. Menezes, T. Kanchan, S.W. Lobo,S. Uysal, N.G.
ce
58.
[73] R. DiBennardo, J.V. Taylor, Multiple discriminant function analysis of sex and
27
Page 31 of 39
[74] S. Mays, M. Cox, Sex determination in skeletal remains, in: M. Cox, S. Mays
t
[76] F. Ramsthaler, K. Kreutz, M.A. Verhoff, Accuracy of metric sex analysis of
ip
skeletal remains using Fordisc based on a recent skull collection, Int. J. Legal
cr
Med. 121 (2007) 477–482.
us
[77] D.H. Ubelaker, Application of Forensic Discriminant Functions to a Spanish
[78]
an
B.A. Williams, T. Rogers, Evaluating the accuracy and precision of cranial
[80] W.M. Bass, Human Osteology: A Laboratory and Field Manual of the Human
pt
[81] W.M. Bass, Human Osteology: A Laboratory and Field Manual of the Human
[82] V. Novotny, M.Y. Iscan, S.R. Loth, Morphologic and osteometric assessment
of age, sex, and race from the skull, in: M.Y. Iscan, R.P. Helmer (Eds.),
Forensic Analysis of the Skull, Wiley-Liss, New York, 1993, pp. 71–88.
(2008) 54-58.
28
Page 32 of 39
[84] T.L. Rogers, Determining the sex of human remains through cranial
t
Forensic Sci. 2 (2012) 81–88.
ip
[86] E. Kranioti, N. Vorniotakis, C. Galiatsou, M.Y. Iscan, M. Michalodimitrakis,
cr
Sex identification and software development using digital femoral head
us
radiographs, Forensic Sci. Int. 189 (1) (2009) 113.e1–7.
[87] P.S. Igbigbi, B.C. Msamati, Sex determination from femoral head diameters in
an
black Malawians. East Afri. Med. J. 77 (3) (2000) 147-151.
[88] S. Black, Forensic osteology in the United Kingdom, in: M. Cox, S. Mays
M
(Eds.), Human Osteology in Archaeology and Forensic Sciences, GMM,
computed tomography (CT) data, J. Med. Assoc. Thai. 93 (1) (2010) 90-98.
29
Page 33 of 39
[92] F.J. Rohlf, Multivariate analysis of shape using partial-warp scores. In: Mardia
t
[94] R.J. Hennessy, C.B. Stringer, Geometric morphometric study of the regional
ip
variation of modern human craniofacial form, Am. J. Phys. Anthropol. 117
cr
(2002) 37–48.
us
[95] A. Rosas, M. Bastir, Thin-plate spline analysis of allometry and sexual
sex estimation of human pelvis, Forensic Sci. Int. 189 (2009) 68–74.
[99] P. Murail, J. Bruzek, F. Houët, E. Cunha, DSP: A tool for probalistic sex
Ac
Page 34 of 39
[101] S. Mestekova, J. Bruzek, J. Veleminska, K. Chaumoitre, A Test of
t
J. Forensic Sci, 48 (3) (2003) 509-512.
ip
[103] M.A. Bidmos, M.R. Dayal, Further evidence to show population specificity of
cr
discriminant function equations for sex determination using the talus of South
us
African Blacks, J. Forensic Sci. 49 (2004) 1–6.
(2005) 129–133.
M
[105] M.Y. Iscan, P. Miller-Shaivitz, Determination of sex from the tibia. Am. J.
[106] B.F.J. Manly, Multivariate Statistical Methods: A Primer, 2 ndedn. Chapman &
[107] F.P. Thieme, W.J. Schull, Sex determination from the skeleton, Hum. Biol. 29
ce
(1957) 242–273.
[109] E. Giles, Discriminant function sexing of the human skeleton, in: T.D. Stewart
[110] T. Waldron, The relative survival of the human skeleton: implication for
Page 35 of 39
Death, Decay and Reconstruction: Approaches to Archaeology and Forensic
t
[112] D.G. Steele, The estimation of sex on the basis of the talus and calcaneus,
ip
Am. J. Phys. Anthropol. 45 (1976) 581–588.
cr
[113] K. Sakaue, Sexual determination of long bones in recent Japanese,
us
Anthropol. Sci. 112 (2004) 75-81.
(2003) 322–328.
[116] Y.C. Luo, Sex determination from the pubis by discriminant function analysis.
pt
39-45.
[118] T.H. Hsiao, H.P. Chang, K.M. Liu, Sex determination by discriminant function
795.
32
Page 36 of 39
[119] R.L. Jantz, S.D. Ousley, FORDISC 1.0–3.0: Personal computerized forensic
t
Forensic Sci. Int. 208 (2011) 180.e1–180.e6.
ip
[121] W.W. Howells, Cranial Variation in Man: A Study by Multivariate Analysis of
cr
Patterns of Difference Among Recent Human Populations, Harvard
us
University, Cambridge: Harvard University Press, 1973.
[122] P. Urbanová, A.H. Ross M. Jurda, M.I. Nogueira, Testing the reliability of
an
software tools in sex and ancestry estimation in a multi-ancestral Brazilian
[124] S.O.Y. Keita, On Meroitic Nubian crania, Fordisc 2.0, and human biological
[126] C.L. Brace, Region does not mean race – reality versus convention in forensic
Ac
[127] G.J. Armelagos, D.P. Van Gerven, A century of skeletal biology and
33
Page 37 of 39
[128] S. Ousley, R. Jantz, D. Freid, Understanding race and human variation: why
evidence in forensic anthropology: sex and race, Am. J. Phys. Anthropol. 139
t
(2009) 77–90.
ip
[130] M.M. Corsini, A. Schmitt, J. Bruzek, Aging process variability on the human
cr
skeleton: artificial network as an appropriate tool for age at death assessment,
us
Forensic Sci. Int. 148 (2005) 163–167.
race in forensic dentistry, Dent. Clin. North Am. 21 (1) (1977) 85–97.
ce
(2015) 96-101.
from the acetabular rim by means of image analysis, Forensic Sci. Int. 180
(2008) 58.e1-e3.
34
Page 38 of 39
[137] A. H. Richard, C. Parks, K. Monson, Accuracy of standard craniometric
measurements using multiple data formats, Forensic Sci. Int. 242 (2014) 177-
185.
[138] M.Y. Iscan, Forensic anthropology of sex and body size, Forensic Sci. Int. 147
(2005) 107–112.
t
[139] K. Krishan, Anthropometry in Forensic Medicine and Forensic Science –
ip
Forensic Anthropometry, Internet J. Forensic Sci. 2007. Available at…
cr
https://ispub.com/IJFS/2/1/10656 (Accessed on 10th January 2015).
us
[140] D. Freid, M.K. Spradley, R.L. Jantz, S.D. Ousley, The truth is out there: how
35
Page 39 of 39