See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/296700791

Body height estimation from post-mortem CT

femoral F1 measurements in a contemporary
Swiss population

Article in Legal Medicine February 2016

DOI: 10.1016/j.legalmed.2016.02.004

CITATIONS READS

3 105

5 authors, including:

Wolf-Dieter Zech Frank Siegmund

Universitt Bern University of Mnster
35 PUBLICATIONS 247 CITATIONS 133 PUBLICATIONS 131 CITATIONS

SEE PROFILE SEE PROFILE

Christian Jackowski Sandra Lsch

Universitt Bern Universitt Bern
111 PUBLICATIONS 2,937 CITATIONS 88 PUBLICATIONS 242 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Forensic Anthropology in Switzerland View project

Nutrition, origin, and social stratification in the Late Iron Age. Stable isotope analysis of Swiss skeletal
series View project

All content following this page was uploaded by Sandra Lsch on 04 March 2016.

The user has requested enhancement of the downloaded file.

 Legal Medicine 19 (2016) 6166

Contents lists available at ScienceDirect

Legal Medicine
journal homepage: www.elsevier.com/locate/legalmed

Body height estimation from post-mortem CT femoral F1 measurements

in a contemporary Swiss population
Wolf-Dieter Zech a,, Maya Nf a, Frank Siegmund b, Christian Jackowski a, Sandra Lsch c
a
Department of Forensic Medicine and Imaging, Institute of Forensic Medicine Bern, University of Bern, Switzerland
b
Institute of History, Heinrich Heine University Dsseldorf, Germany
c
Department of Physical Anthropology, Institute of Forensic Medicine, University of Bern, Switzerland

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

Article history: Purpose: The present study aimed at the comparison of body height estimations from cadaver length
Received 28 November 2015 with body height estimations according to Trotter and Gleser (1952) and Penning and Riepert (2003)
Received in revised form 15 February 2016 on the basis of femoral F1 section measurements in post-mortem computed tomography (PMCT) images.
Accepted 16 February 2016
Methods: In a post-mortem study in a contemporary Swiss population (226 corpses: 143 males (mean
Available online 16 February 2016
age: 53 17 years) and 83 females (mean age: 61 20 years)) femoral F1 measurements (403 femora:
199 right and 204 left; 177 pairs) were conducted in PMCT images and F1 was used for body height esti-
Keywords:
mation using the equations after Trotter and Gleser (1952, American Whites), and Penning and Riepert
Stature
Forensic anthropology
(2003).
Post-mortem computed tomography Results: The mean observed cadaver length was 176.6 cm in males and 163.6 cm in females. Mean mea-
(PMCT) sured femoral length F1 was 47.5 cm (males) and 44.1 cm (females) respectively. Comparison of body
Femur height estimated from PMCT F1 measurements with body height calculated from cadaver length showed
Body height estimation a close congruence (mean difference less than 0.95 cm in males and less than 1.99 cm in females) for
equations both applied after Penning and Riepert and Trotter and Gleser.
Conclusions: Femoral F1 measurements in PMCT images are very accurate, reproducible and feasible for
body height estimation of a contemporary Swiss population when using the equations after Penning and
Riepert (2003) or Trotter and Gleser (1952).
2016 Elsevier Ireland Ltd. All rights reserved.

1. Introduction
Over the last 10 years post-mortem CT (PMCT) scanning has
become a useful tool in forensic practice [13]. Since CT images
can depict osseous structures well, CT imaging is also applied in
forensic anthropology for tasks such as determination of sex, age
and body height [49]. Body height estimation remains one of
the key parameters to identify unknown individuals [10,11]. Body
height can be estimated by the anatomical method or mathemat-
ically by the use of regression equations. The anatomical method
provides more accurate estimates, but requires the entire skeleton
by summing up measuring lines from the skull to the foot [12,13].
In forensic practice decomposed, burned or incompletely pre-
served corpses may be presented as the only available source to
establish identity. In such cases measurement of single skeletal
Corresponding author at: Department of Forensic Medicine and Imaging,
Institute of Forensic Medicine Bern, University of Bern, Bhlstrasse 20, 3012 CH
Bern, Switzerland.
E-mail address: wolf-dieter.zech@irm.unibe.ch (W.-D. Zech).
elements and their linear proportionality to the stature can be used
for body height estimation or estimation of sex. In settings, such as
mass disasters the use of PMCT imaging can be very helpful
because metric bone measurements of any body part can be exe-
cuted quickly, are easily to be reproduced and avoid elaborate
maceration procedures [3,6,8,9,1417]. Moreover CT data offer
the advantage of population studies and enable for bone recon-
struction as a means of providing metrical data of skeletal struc-
tures in victim identification. CT measurements may also be
conducted for determination of sex [18,19]. Several previous stud-
ies investigated the feasibility of CT bone measurements such as
pelvic measurements or radial and ulnar length measurements
for body height estimation in Japanese and Chinese populations
[2023]. Other studies used the metatarsal bones in a Spanish pop-
ulation or the first cervical bones in a Portuguese population for
body height estimation from CT measurements [18,19]. Since
femur length is closely correlated with stature, metric femoral
parameters can be used to determine the body height of an individ-
ual [2431]. Hence PMCT measurements of metric femoral param-
eters can be used to determine body height as well [23,32]. Several

http://dx.doi.org/10.1016/j.legalmed.2016.02.004
1344-6223/ 2016 Elsevier Ireland Ltd. All rights reserved.
62 W.-D. Zech et al. / Legal Medicine 19 (2016) 6166
equations for body height estimation from the femur have already
been established. Only few of them have been re-evaluated since
[11,13,33,34]. Trotter and Gleser [35] developed different equa-
tions for White, Negroid and Mongoloid populations. The
equations generated by Penning and Riepert [36] were established
on a large southern German population sample. However, evalua-
tion of these equations has shown regional and temporal bias [11].
The present study aimed at the comparison of body height estima-
tions from cadaver length with body height estimations according
to Trotter and Gleser [35] and Penning and Riepert [36] on the
basis of measurements of the anthropologically standardized
femoral F1 section in PMCT images. Furthermore, accuracy of body
height estimation between the regression equations after Trotter
and Gleser [35] and the regression equations after Penning and
Riepert [36] for a contemporary Swiss population sample should
be assessed.
2. Materials and methods
2.1. Study subjects
The study was performed retrospectively on PMCT images of
226 corpses (143 males and 83 females). Information about the liv-
ing height of the individuals e.g. from passport or medical docu-
ments was not available. The sample consisted of total 403
femora (199 right and 204 left, thereof 177 pairs). Study cases were
chosen by the following criteria (as stated in the autopsy reports):
(1) confirmed age between 20 and 95 years, (2) skeletal system: no
fractures of lower extremities, no fractures of vertebral column, no
severe head fractures, (3) Caucasian origin, (4) no signs of relevant
decomposition according to autopsy reports. In cases with one
sided hip replacements and/or knee replacements the counterpart
femur was measured. The mean age of the men was 53 17 years
and that of the women 61 20 years. The post mortem interval
between death and CT scan/autopsy ranged from one day to
10 days.
2.2. Measurement of cadaver length and estimation of body height
After PMCT scanning the cadaver length (CL) was measured in
whole centimeters using a stiff yardstick. CL was defined as maxi-
mum distance between skull vertex and the sole of the heels in 90
angle to the anterior part of the lower limbs. Bodies were
unclothed and in supine position on a plane metal table. According
to Trotter and Gleser [35] measured cadaver length was converted
into body height (BH/CL) by subtracting 2.5 cm. The mean, stan-
dard deviation and interquartile range of measured cadaver length
were calculated for all subjects. KolmogorovSmirnov test (KS
test) for goodness of fit was used to test for normality of the distri-
bution of cadaver length and femur length within the study popu-
lation [37].
2.3. PMCT imaging and image analysis
All corpses underwent PMCT scan (6-slice Somatom Emotion 6,
Siemens Medical Solution, Forchheim, Germany) prior to forensic
autopsy. The interval between CT scan and forensic autopsy was
maximum one day. The scan parameters for raw data acquisition
were as follows: beam energy 130 kilovolt (kV); rotation time
1500 ms. Image reconstruction was performed as follows: kernel
B70; slice thickness 1.25 mm; increment: 0.7 mm; the field of view
was adapted to the size of the object.
Measurements in PMCT images were conducted on a Leonardo
Workstation (Siemens, Forchheim, Germany) by one observer.
PMCT images of the femora were examined in multi-planar mode.
According to Martin [38] the metric F1 parameter (distance
between the highest point of the caput femoris and the lowest
point of the medial condyle on a conventional osteometric board)
was measured in PMCT images (Fig. 1). The mean, standard devia-
tion and interquartile range of measured femoral length F1 in
PMCT images were calculated for all subjects. When both femora
were measured in the same individual, the arithmetic mean of
both measurements was used. KolmogorovSmirnov goodness of
fit test was used to test for normality of the distribution of F1 mea-
surements [37].
2.4. Intra-observer error
Intra-observer error of PMCT F1 measurements in corpses of the
study population was estimated using randomly selected PMCT
images of 20 right femora (of 10 male and 10 female bodies). After
a two month period F1 was measured again in the same PMCT
images by the same observer. The intra-observer error was esti-
mated calculating the difference and the absolute difference
between the two measurement series of F1. A t-test for paired sam-
ples was applied to test the significance of the differences [37].
2.5. Inter-observer error
Inter-observer error of PMCT F1 measurements in corpses of the
study population was estimated using randomly selected PMCT
images of 20 right femora (of 10 male and 10 female bodies).
Two observers measured F1 of the same individuals in PMCT
images. The inter-observer error was estimated calculating the dif-
ference and the absolute difference between the measurements of
the two observers. A t-test for paired samples was applied to test
the significance of the differences [37].
2.6. Accuracy of PMCT F1 measurements vs osteometric board
To test the accuracy of F1 measurements taken from PMCT, five
observers independently measured F1 values at archeological
femoral bones (8 femora per observer: 4 left, 4 right) with a con-
ventional osteometric board. Afterwards the same femora were
CT-scanned with their dorsal side on the plane CT examination
table and F1 was measured in PMCT images by the same observers
(Fig. 2). The arithmetic mean of all F1 measurements conducted
Fig. 1. F1 measurements in PMCT images: The Femur was positioned in the sagittal
plane. The coronal axis (green line) was adjusted to the most dorsal points
(Trochanter major and condyles) of the Femur (a). By adjusting the sagittal plane,
the highest point of the Caput was determined and marked with a line (blue axial
axis) (b). At the coronal plane, the sagittal axis (red line) was drawn between the
lowest point of the medial condyle and the highest point of the caput femoris (c).
Then, F1 was vertically measured in the sagittal oblique plane between the lowest
point of the medial condyle and the highest point of the caput femoris (d). (For
interpretation of the references to colour in this figure legend, the reader is referred
to the web version of this article.)
 W.-D. Zech et al. / Legal Medicine 19 (2016) 6166
Fig. 2. To test for accuracy of PMCT F1 measurements F1 was measured on
defleshed femoral bones on a conventional osteometric board (a) and afterwards in
CT images (b).
with the osteometric board was calculated and defined as real
femoral length (RFL). Afterwards the differences (d) of the real
femoral length and each single measured PMCT F1 value were cal-
culated for each observer. Significant differences of d between the
five observers were estimated using the KruskalWallis-H-Test.
Significant differences between the observations on the osteomet-
ric board and the PMCT-Images were estimated using the t-test for
paired samples [37].
2.7. Estimation of body height from PMCT F1 measurements
Body height of the study subjects which underwent PMCT scans
was estimated on the basis of the F1 data from femoral measure-
ments in PMCT images (BH/F1). Estimations were performed by
means of regression equations by Penning and Riepert [36] and
Trotter and Gleser [35] (see Supplementary material: estimation
equations). Equations were applied with and without age
correction.
2.8. Comparison between BH/F1 and BH/CL
The mean differences between BH/CL and BH/F1 were calcu-
lated. Calculations were applied for BH/F1 with and without age
correction according to the regression equations by Penning and
Riepert [36] and Trotter and Gleser [35]. A t-test for paired samples
was applied for the evaluation of significant differences between
BH/CL and BH/F1 values.
2.9. Statistical analysis
Statistical analyses were performed using SPSS version 22
computer software (IBM Corp., Armonk, NY).
Table 1
Mean cadaver length (CL, in cm) of the study population. IQR = interquartile range. Cadaver length is not significantly deviant from normal distribution in both sexes (KS test).
CL mean SD
Males (n = 143) 176.6 7.2
females (n = 83) 163.6 7.8
63
3. Results
3.1. Cadaver length and femur length F1 of the population
Table 1 shows the mean of the individuals cadaver length (CL).
Table 2 shows the mean femoral F1 values of the population mea-
sured in PMCT images. CL and F1 were normally distributed in both
sexes (KS test). Therefore, parametric statistics and tests were
used.
3.2. Intra-observer error
The measurements for intra-observer error at the PMCT images
showed a mean difference between the first and the second mea-
surement of 0.5 mm, indicating that there is no relevant differ-
ence between the two series of measurements. The absolute
difference between the first and second measured value of F1
was 0.7 mm (Table 3). A t-test for paired values shows that the
small differences between the first and the second measurement
are statistically significant (t 2.932, sign. 0.009). Applied to the
material and estimation methods, a difference of 0.5 mm at F1
would cause a difference in body height estimation of approx.
1 mm.
3.3. Inter-observer error
The measurements for inter-observer error at the PMCT images
showed a mean difference between the first and the second obser-
ver of 0.4 mm, indicating that there is no relevant difference
between the two series of measurements. The absolute difference
between the first and second observer was 1.3 mm (Table 4). A t-
test for paired values confirms that the differences between the
two observers are not significant (t 0.709, sign. 0.487). Applied
to the material and estimation methods, a difference of 1.3 mm
at F1 would cause a difference in body height estimation of approx.
3 mm.
3.4. Accuracy of PMCT F1 measurements vs osteometric board
The F1 measurements of 8 archaeological femora on an osteo-
metric board showed an arithmetic mean of 465.2 mm 32.8
(real femoral length). There were no significant differences of
the measured PMCT F1 values between the five different observers
(KruskalWallis-H-Test: chi-square 2.727, significance 0.604).
Compared to the osteometric board the PMCT measurements
exhibited a slightly decreased F1 mean value of 0.6 mm with a
slightly increased standard deviation (Table 5). T-test for paired
samples between F1 measurements conducted on the 8 archaeo-
logical femora (with an osteometric board) compared to the F1
measurements obtained from PMCT images of the same bones
showed, that these differences although small were statistically
significant (t = 3.041, significance 0.004). Applied to the material
and estimation methods, a difference of 0.6 mm at F1 would cause
a difference in body height estimation of approx. 1.5 mm.
Minmax Median IQR KS test
158199 176.0 173181 0.072, sign. 0.433
143182 163.0 159169 0.059, sign. 0.920
64 W.-D. Zech et al. / Legal Medicine 19 (2016) 6166

Table 2
Femoral length F1 (in cm) of the study population. IQR = interquartile range. F1 is not significantly deviant from normal distribution in both sexes (KS test).

F1 mean SD Minmax Median IQR KS test

Males (n = 143) 47.5 2.19 41.753.4 47.6 46.049.2 0.042, sign. 0.957
Females (n = 83) 44.1 2.37 38.950.3 44.0 42.445.8 0.035, sign. 1.000

Table 3
Results for intra-observer error estimation (n = 20). The mean difference between the first and second measured value of F1 was 0.5 mm (column 2), and the mean absolute
difference was 0.7 mm (column 4); the maximum aberration was 1 mm (column 5).

d mean SD (mm) Minmax (mm) d absolute mean SD (mm) Minmax (mm)

PMCT F1 0.5 0.7 1 to 1 0.7 0.5 01

Table 4
Results for inter-observer error estimation (n = 20). The mean difference between the first and second observer measuring F1 was 0.4 mm (column 2), and the mean absolute
difference was 1.3 mm (column 4); the maximum aberration was 6 mm (column 5).

d mean SD (mm) Minmax (mm) d absolute mean SD (mm) Minmax (mm)

PMCT F1 0.4 2.2 6 to 5 1.3 1.8 06

Table 5
Accuracy of F1 measurements in CT images of archeological femoral bones compared to osteometric board F1 measurements (n = 40): The mean difference (board minus CT) was
+0.6 mm (column 2) and the mean absolute difference was 1.1 mm (column 4), the maximum aberration was 3 mm (column 5).

d mean SD (mm) Minmax (mm) d absolute mean SD (mm) Minmax (mm)

BoardCT 0.6 1.2 3 to 3 1.1 0.8 03

Table 6
Relation between the mean body height calculated from measurements of cadaver length and the mean body height estimated from PMCT F1 measurements after Trotter/Gleser
(1952) and Penning/Riepert (2003) for the males (n = 143). Difference = estimated BH/F1 minus calculated BH/CL. For all estimation equations, the difference between BH/CL and
BH/F1 is not significant (t-test for paired samples). All values are in centimeter with standard deviation.

Mean body height (cm) estimated from F1 Mean difference (cm) to BH

measurements in PMCT (BH/F1) calculated from CL (BH/CL)
Trotter/Gleser 1952 white, not age corrected 174.5 5.2 +0.47 4.95
Trotter/Gleser 1952 white, age corrected 173.1 5.5 0.95 4.68
Penning/Riepert 2003, without age 173.8 5.8 0.26 4.98
Penning/Riepert 2003, 10-years-model 174.2 6.2 +0.12 4.64
Penning/Riepert 2003, 20-years-model 173.3 6.1 0.77 4.67

Table 7
Relation between the mean body height calculated from measurements of cadaver length and the mean body height estimated from PMCT F1 measurements after Trotter/Gleser
(1952) and Penning/Riepert (2003) for the females (n = 83). Difference = estimated BH/F1 minus calculated BH/CL. For all estimation equations with age correction, the difference
between BH/CL and BH/F1 is not significant (t-test for paired samples); for both estimation equations without age correction, the small differences between BH/CL and BH/F1 are
significant (t-test for paired samples). All values are in centimeter with standard deviation.

Mean body height (cm) estimated from F1 Mean difference (cm) to BH

measurements in PMCT (BH/F1) calculated from CL (BH/CL)
Trotter/Gleser 1952 white, not age corrected 163.0 5.8 +1.98 4.81
Trotter/Gleser 1952 white, age corrected 161.2 6.2 +0.12 4.36
Penning/Riepert 2003, without age 162.7 5.8 +1.62 4.82
Penning/Riepert 2003, 10-years-model 161.5 6.3 +0.39 4.41
Penning/Riepert 2003, 20-years-model 161.3 5.9 +0.26 4.51

3.5. Comparison of body height estimated from PMCT F1
measurements (BH/F1) with cadaver length measurements (BH/CL)
Tables 6 and 7 show the mean difference between body height
calculated from the observed cadaver length (i.e. BH/CL = CL minus
2.5 cm; males 174.1 7.2, females 161.1 7.8 cm) and the esti-
mated BH from F1 after the five regression approaches (BH/F1).
For males (Table 6), all estimations (BH/F1) are close to the calcu-
lated body height (BH/CL), with deviations lesser than 1 cm. A t-
test for paired samples indicates that these differences are not sig-
nificant. Both equations without age correction show smaller devi-
ations of less than a half centimeter, while Trotter and Gleser [35]
with age correction and the 20-years-model of Penning and Riepert
[36] show higher deviations. For females (Table 7), both estima-
tions without age correction show an overestimation of body
height of 1.98 resp. 1.62 cm, these differences are statistically sig-
nificant (t-test for paired samples). On the contrary, the estima-
tions with age correction show very small deviations from
 W.-D. Zech et al. / Legal Medicine 19 (2016) 6166
calculated body height (BH/CL), with differences less than a half
centimeter.
4. Discussion
In the present study, metric data F1 data of the femur were
obtained from post-mortem CT images for estimation of body
height. The study proved that the applied metric femoral F1 mea-
surements in PMCT images were very accurate and reproducible
between different observers. There is a small, but statistically sig-
nificant difference between manual measurements of F1 on
archaeological femoral bones on a conventional osteometric board
compared to F1 measurements in PMCT. When applied to the esti-
mation equations used here, this systematic difference of ca.
0.6 mm would cause an underestimation of body height of about
1.5 mm by PMCT images in comparison to the osteometric board,
a difference, which could be neglected. Therefore, it can be con-
cluded that PMCT femoral F1 measurements as conducted in this
study provide reproducible and valid metric results which can be
used for body height estimation with appropriate regression equa-
tions. It is noteworthy determination of height in corpses in gen-
eral is not limited to the use of femoral measurements. Forensic
anthropologists rely on several other bones of the human body
for body height estimation. Several studies for the determination
of body height from bone measurements in CT images have been
conducted not for the femur but several other bones of the human
body such as scapula, forearm bones, first and second cervical ver-
tebra or metatarsal bones. It had been shown that metric measure-
ments of these bones are feasible for the determination of body
height. In many cases measurements of such smaller bones may
even be more feasible since long bones are often missing or frag-
mented [1824]. However the results of the present study as well
as previous studies demonstrate that femoral measurements in CT
images can play an important role in body height estimation of
corpses or skeletal remains [23]. Moreover, the usage of CT imaging
data bears the advantage for elimination of soft tissue surrounding
the bone which can make CT measurements even more accurate
than direct bone measurements [19]. As demonstrated by Rodri-
guez et al. the method of highly accurate bone measurements in
CT images may also be conducted on CT images of living persons
and being used for contemporary population studies [18].
Since there are globally observed changes in secularization, the
issue of outdated regression equations for body height estimation
has repeatedly been raised [11,13,15,31,38,39]. Equations cali-
brated for traits such as body height or sex are usually applicable
to subjects from the same ethnic group [11,15,27,39]. This is one
of the reasons why recent research partially concentrated on
development of new regression equations from femoral measure-
ments for selected populations [40,41]. However, few of the exist-
ing equations for body height estimation using femoral
measurements have been re-evaluated for other populations.
Therefore, in the present study instead of developing new equa-
tions, the already existing equations after Penning and Riepert
[36] and Trotter and Gleser [35] which were developed on a Ger-
man population and on north American populations respectively
were evaluated for their feasibility for a contemporary Swiss pop-
ulation [35,36]. The body height estimations according to Penning
and Riepert [36] and Trotter and Gleser [35] on the basis of the F1
measurements conducted in PMCT images led to very similar and
comparable results. Thus, on the one hand the equations after Trot-
ter and Gleser [35] could be confirmed to be accurate for the con-
temporary Swiss population. On the other hand the lesser known
equations after Penning and Riepert [36] could be validated on this
contemporary Swiss population as well. The results indicate that if
possible the estimations with age correction should be preferred.
65
Imaging methods (e.g. ultrasound measurement in order to
assess fetal growth) have been used for anthropometry since the
late 1950s already [42]. Recently in vivo CT and MRI studies con-
ducted femoral bone measurements in CT and MRI images in order
to determine bone lengths and to develop new regression equa-
tions for body height estimation in different populations
[40,43,44]. However none of these studies manually validated the
accuracy of their measurement techniques for femoral bone
lengths. It has to be taken into account that accuracy and repro-
ducibility of bone measurements not only depends on image qual-
ity but also on precise definition of landmarks [2530]. The results
from the present study showed that the PMCT F1 measurement
technique as conducted in the present study was as accurate as
measurements on a conventional osteometric board. The so far
conducted in vivo CT and MRI studies which used femoral length
for body height estimation did not use anthropologically standard-
ized measurement sections [40,43,44]. Therefore inaccurate bone
measurements may be the source of errors in calibrating regres-
sion equations and estimating body height. In the present study
the metric parameter F1 after Martin [38] as clearly defined mea-
surement section was measured in PMCT images and showed a
close congruence with body height. Recent work from Giurazza
et al. and Hishmat et al. conducted femoral CT measurements
without using standardized anthropological sections of the femur
with results that showed a strong correlation between body height
and femoral measurements [23,40]. This indicates that body height
estimation from femoral measurements in CT images is also feasi-
ble when using none anthropologically standardized measurement
sections. However Hishmat et al. conducted measurements on a
post-mortem Japanese population and Giurazza et al. conducted
measurements on living persons exclusively which limits the com-
parability with these studies. In comparison to the other studies
the PMCT F1 femoral bone measurement in the present study is
more elaborate than measurement of anthropologically not stan-
dardized sections. However the F1 measurements as conducted
in this study are relatively easy to perform with a common radio-
logical workstation and can be assumed to be accurate.
4.1. Limitations of the study
Since cadaveric measurements are known to be imprecise the
corrected body heights evaluated from cadaver length remain an
uncertainty factor [41]. Other studies however found no significant
difference between cadaver length and previous heights records
[45]. In the present study 2.5 cm were subtracted from cadaver
length to assess real body height according to Trotter and Gleser
[35] which led to results that showed a close congruence between
body height estimated from PMCT F1 measurements and body
height calculated from cadaver length. However other recent
works suggest that this correction should not be applied [46].
It has to be taken into account that slice thickness of CT images
was 1.25 mm. Due to the thickness of the first and last slice of the
measured femora an error range from 1.25 mm to 2.5 mm may be
considered. Such error range was not considered for calculations of
body height from F1 measurements. However the body height esti-
mations in the present study on the basis of the conducted F1 mea-
surements led to very similar and comparable results. Therefore
errors in F1 resulting from CT slice thickness may be neglected
when used for body height estimation.
5. Conclusion
It can be concluded that femoral F1 measurements in PMCT
images are very accurate, reproducible and feasible for body height
estimation of a contemporary Swiss population when using the
66 W.-D. Zech et al. / Legal Medicine 19 (2016) 6166
equations after Penning and Riepert (2003) or Trotter and Gleser
(1952). In the contemporary Swiss population sample of this study
the equations after Penning and Riepert [36] are as good for body
height estimation as the equations after Trotter and Gleser [35].
Acknowledgements
We would like to thank Simon Kramis for his valuable help,
Negahnaz Moghaddam and Domenic Rttimann for their help with
manual bone measurements. We are grateful to Amelie Alterauge
for help with the literature-management. We also thank our team
of forensic technicians and forensic pathologists for their support
and case handling.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.legalmed.2016.
02.004.
References
[1] A. Persson, C. Jackowski, E. Engstrom, H. Zachrisson, Advances of dual source,
dual-energy imaging in postmortem CT, Eur. J. Radiol. 68 (2008) 446455.
[2] W.D. Zech, C. Jackowski, Y. Buetikofer, L. Kara, Characterization and
differentiation of body fluids, putrefaction fluid, and blood using Hounsfield
unit in postmortem CT, Int. J. Legal Med. 128 (2014) 795802.
[3] M.J. Thali, K. Yen, W. Schweitzer, P. Vock, C. Boesch, C. Ozdoba, et al., Virtopsy,
a new imaging horizon in forensic pathology: Virtual autopsy by postmortem
multislice computed tomography (MSCT) and magnetic resonance imaging
(MRI) -a feasibility study, J. Forensic Sci. 48 (2003) 386403.
[4] F. Dedouit, N. Telmon, R. Costagliola, P. Otal, F. Joffre, D. Rouge, Virtual
anthropology and forensic identification: report of one case, Forensic Sci. Int.
173 (2007) 182187.
[5] S. Schmidt, I. Nitz, R. Schulz, A. Schmeling, Applicability of the skeletal age
determination method of Tanner and Whitehouse for forensic age diagnostics,
Int. J. Legal Med. 122 (2008) 309314.
[6] S. Grabherr, C. Cooper, S. Ulrich-Bochsler, T. Uldin, S. Ross, L. Oesterhelweg,
et al., Estimation of sex and age of virtual skeletons a feasibility study, Eur.
Radiol. 19 (2009) 419429.
[7] F. Ramsthaler, M. Kettner, A. Gehl, M.A. Verhoff, Digital forensic osteology:
Morphological sexing of skeletal remains using volume-rendered cranial CT
scans, Forensic Sci. Int. 195 (2010) 148152.
[8] W.D. Zech, G. Hatch, L. Siegenthaler, M.J. Thali, S. Lsch, Sex determination
from os sacrum by postmortem CT, Forensic Sci. Int. 221 (2012) 3943.
[9] C. ODonnell, M. Iino, K. Mansharan, J. Leditscke, N. Woodford, Contribution of
postmortem multi detector CT scanning to identification of the deceased in a
mass disaster: experience gained from the 2009 Victorian bushfires, Forensic
Sci. Int. 205 (2011) 1528.
[10] B. Knight, Forensic Pathology, Second ed., CRC Press, Boca Raton, 1996.
[11] R.J. Wilson, N.P. Herrmann, L.M. Jantz, Evaluation of stature estimation from
the database for forensic anthropology, J. Forensic Sci. 55 (2010) 684689.
[12] M.A. Bidmos, Metatarsals in the estimation of stature in South Africans, J.
Forensic Leg. Med. 15 (2008) 505509.
[13] M.H. Raxter, B.M. Auerbach, C.B. Ruff, Revision of the fully technique for
estimating statures, Am. J. Phys. Anthropol. 130 (2006) 374384.
[14] M. Cox, S. Mays, Human Osteology in Archaeology and Forensic Science,
Greenwich Medical Media, London, 2000.
[15] W.M. Krogman, M.Y. Is _can, The Human Skeleton in Forensic Medicine, C.C.
Thomas, Springfield IL, 1986.
[16] M. Muthusubramanian, K.S. Limson, R. Julian, Analysis of rugae in burn victims
and cadavers to simulate rugae identification in cases of incineration and
decomposition, J. Forensic Odontostomatol. 23 (2005).
[17] T. Delabarde, B. Ludes, Missing in Amazonian jungle: a case report of skeletal
evidence for dismemberment, J. Forensic Sci. 55 (2010) 11051110.
[18] S. Rodrguez, A. Gonzlez, A. Simn, M.S. Rodrguez-Calvo, M. Febrero-Bande,
C. Cordeiro, J.I. Muoz-Bars, The use of computerized tomography in
determining stature and sex from metatarsal bones, Leg Med. (Tokyo) 16 (5)
(2014) 252257.
[19] S. Rodrguez, M.S. Rodrguez-Calvo, A. Gonzlez, M. Febrero-Bande, J.I. Muoz-
Bars, Estimating height from the first and second cervical vertebrae in a
Spanish population, Leg Med. (Tokyo) (2015). pii: S1344-6223(15)30017-1.
[20] S. Torimitsu, Y. Makino, H. Saitoh, A. Sakuma, N. Ishii, M. Hayakawa, D. Yajima,
G. Inokuchi, A. Motomura, F. Chiba, H. Iwase, Stature estimation based on
View publication stats
radial and ulnar lengths using three-dimensional images from multidetector
computed tomography in a Japanese population, Leg Med. (Tokyo) 16 (4)
(2014) 181186.
[21] S. Torimitsu, Y. Makino, H. Saitoh, A. Sakuma, N. Ishii, M. Hayakawa, D. Yajima,
G. Inokuchi, A. Motomura, F. Chiba, H. Iwase, Stature estimation in Japanese
cadavers based on pelvic measurements in three-dimensional multidetector
computed tomographic images, Int. J. Legal Med. 129 (3) (2015) 633639.
[22] K. Zhang, Y.Z. Luo, F. Fan, J.Q. Zheng, M. Yang, T. Li, T. Pang, J. Zhang, Z.H. Deng,
Stature estimation from sternum length using computed tomography-volume
rendering technique images of Western Chinese, J. Forensic Leg. Med. 35
(2015) 4044.
[23] A.M. Hishmat, T. Michiue, N. Sogawa, S. Oritani, T. Ishikawa, I.A. Fawzy, M.A.
Hashem, H. Maeda, Virtual CT morphometry of lower limb long bones for
estimation of the sex and stature using postmortem Japanese adult data in
forensic identification, Int. J. Legal Med. 129 (5) (2015) 11731182.
[24] S. Rodrguez, X. Miguns, M.S. Rodrguez-Calvo, M. Febrero-Bande, J.I. Muoz-
Bars, Estimating adult stature from radiographically determined metatarsal
length in a Spanish population, Forensic Sci. Int. 226 (13) (2013) 297.e1-4.
[25] D. Lorke, D. Mnzer, E. Walter, Zur Rekonstruktion der Krpergrsse eines
Menschen aus den langen Gliedmassenknochen, Deutsche Zeitschrift fr
gerichtliche Medizin 42 (1953) 189202.
[26] S. Genoves, Proportionality of long bones and their relation to stature among
Mesoamericans, Am. J. Phys. Anthropol. 26 (1967). 67-&.
[27] C.W. Dupertuis, J.A. Hadden, On the reconstruction of stature from long bones,
Am. J. Phys. Anthropol. 9 (1951) 1553.
[28] E. Breitinger, Zur Berechnung der Krperhhe aus den langen
Gliedmaenknochen, Anthropol. Anz. 14 (1937) 249274.
[29] H. Bach, Zur Berechnung der Krperhhe aus den langen Gliedmaenknochen
weiblicher Skelette, Anthropol. Anz. 29 (1965) 1221.
[30] E. Sonder, R. Knumann, Zur Krperhhenbestimmung mnnlicher Individuen
aus Femur-, Tibia- und Humerus-Fragmenten: on the determination of the
stature of male individuals using femoral, tibial, and humeral fragments, Z.
Morphol. Anthropol. 75 (1985) 131153.
[31] F. Siegmund, Die Krpergrsse der Menschen in der Ur- und Frhgeschichte
Mitteleuropas und ein Vergleich ihrer anthropologischen Schtzmethoden,
Books on Demand GmbH, Norderstedt, 2010.
[32] F. Giurazza, R. Del Vescovo, E. Schena, R.L. Cazzato, F. DAgostino, R.F. Grasso,
et al., Stature estimation from scapular measurements by CT scan evaluation
in an Italian population, Leg. Med. 15 (2013) 202208.
[33] T.P. Gocha, G. Vercellotti, L.E. McCormick, T.L. Van Deest, Formulae for
estimating skeletal height in modern South-East Asians, J. Forensic Sci. 58
(2013) 12791283.
[34] B.R. Kate, R.D. Mujumdar, Stature estimation from femur and humerus by
regression and autometry, Acta Anatomica (Basel) 94 (1976) 311320.
[35] M. Trotter, G.C. Gleser, Estimation of stature from long bones of American
Whites and Negroes, Am. J. Phys. Anthropol. 10 (1952) 463514.
[36] R. Penning, T. Riepert, Krpergrenschtzformeln aus Langknochenmaen,
in: B. Madea, B. Brinkmann (Eds.), Handbuch gerichtliche Medizin auf DVD,
Springer, Berlin, Heidelberg, New York, Hongkong, London, Mailand, Paris,
Tokio, 2003, pp. 11261133.
[37] R.R. Sokal, F.J. Rohlf, Biometry: The Principles and Practices of Statistics in
Biological Research, 3rd ed., W.H. Freeman, New York, 1994.
[38] R. Martin, Lehrbuch der Anthropologie in systematischer Darstellung mit
besonderer Bercksichtigung der anthropologischen Methoden, Gustav
Fischer, Jena, 1914.
[39] M.R. Dayal, M.A. Spocter, M.A. Bidmos, An assessment of sex using the skull of
black South Africans by discriminant function analysis, HOMO 59 (2008) 209
221.
[40] F. Giurazza, R. Del Vescovo, E. Schena, S. Battisti, R.L. Cazzato, F.R. Grasso, et al.,
Determination of stature from skeletal and skull measurements by CT scan
evaluation, Forensic Sci. Int. 222 (2012). 398.e1-.e9.
[41] H. Karakas, O. Celbis, A. Harma, B. Alicioglu, Total body height estimation using
sacrum height in Anatolian Caucasians: multidetector computed tomography-
based virtual anthropometry, Skeletal Radiol. 40 (2011) 623630.
[42] R.L. Deter, R.B. Harrist, F.P. Hadlock, R.J. Carpenter, The use of ultrasound in the
assessment of normal fetal growth: a review, J. Clin. Ultrasound 9 (1981) 481
493.
[43] M.A. Bidmos, P.R. Manger, New soft tissue correction factors for stature
estimation: results from magnetic resonance imaging, Forensic Sci. Int. 214
(2012). 212.e1-.e7.
[44] I. Hasegawa, K. Uenishi, T. Fukunaga, R. Kimura, M. Osawa, Stature estimation
formulae from radiographically determined limb bone length in a modern
Japanese population, Leg. Med. 11 (2009) 260266.
[45] C. Cordeiro, J.I. Munoz-Bars, S. Wasterlain, E. Cunha, D.N. Vieira, Predicting
adult stature from metatarsal length in a Portuguese population, Forensic Sci.
Int. 193 (2009). 131.e1.e4.
[46] A. Pablos, A. Gmez-Olivencia, A. Garca-Prez, I. Martnez, C. Lorenzo, J.L.
Arsuaga, From toe to head: use of robust regression methods in stature
estimation based on foot remains, Forensic Sci. Int. 226 (2013) 299. 299.e1.
e7.