Journal of Forensic and Legal Medicine 35 (2015) 40e44

Contents lists available at ScienceDirect

Journal of Forensic and Legal Medicine

j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / j fl m

Original communication

Stature estimation from sternum length using computed

tomographyevolume rendering technique images of western Chinese
Kui Zhang a, g, Ying-zhen Luo a, g, Fei Fan a, Jie-qian Zheng b, Min Yang c, Tao Li d,
Tao Pang e, Jian Zhang f, Zhen-hua Deng a, *
a
Department of Forensic Pathology, West China School of Preclinical and Forensic Medicine, Sichuan University, Chengdu, Sichuan 610041, PR China
b
Department of Radiology, West China Hospital, Sichuan University, Chengdu, Sichuan 610041, PR China
c
Nan Chong Public Security Bureau, Nan Chong, Sichuan 610000, PR China
d
Peng An Public Security Bureau, Nan Chong, Sichuan 610000, PR China
e
Gao Ping Public Security Bureau, Nan Chong, Sichuan 610000, PR China
f
Xi Chong Public Security Bureau, Nan Chong, Sichuan 610000, PR China

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

Article history: The objective of the present investigation was to generate linear regression models for stature estimation
Received 20 January 2015 on the basis of sternum length derived from computed tomographyevolume rendering technique (CT
Received in revised form eVRT) images for Western Chinese. The study sample comprised 288 individuals of Western Chinese,
20 June 2015
including 124 females and 164 males, with documented ages between 19 and 78 years, and was
Accepted 23 July 2015
Available online 1 August 2015
randomly divided into two subgroups. The linear regression analysis for the calibration sample data
yielded the following formulae: male stature (cm) ¼ 137.28 þ 1.99*combined length of manubrium and
mesosternum and female stature (cm) ¼ 111.59 þ 3.51* combined length of manubrium and meso-
Keywords:
Forensic science
sternum. Pearson's correlation coefficients for the regression models were r ¼ 0.459 and r ¼ 0.541 for the
Forensic anthropology male and female formulae, respectively. The standard errors of the estimate (SEE) were 4.76 cm for the
CTeVRT male equation and 6.73 cm for the female equation. The 95% confidence intervals of the predicted values
Linear regression encompassed the correct stature of all specimen in the validation sample. The regression equations
derived from the sternum length in the present study can be used for stature estimation and the length
of the sternum is a reliable predictor of stature in Chinese when better predictors of stature like the long
bones are not available, and the CTeVRT method may be a practical method for stature estimation.
© 2015 Elsevier Ltd and Faculty of Forensic and Legal Medicine. All rights reserved.

1. Introduction elements have been used for the derivation of regression equa-
tions for the estimation of stature, such as long bones,3e8 scap-
Forensic anthropology plays a crucial role in the medico-legal ula,9e11 metacarpals,12 metatarsals,13 and vertebral column.14e17
investigation of unidentified skeletonized or decomposed human Recently, regression equations for stature estimation based on
remains. The estimation of stature is an important component. The the length of the sternum have been developed from the mea-
most accurate method for reconstructing stature involves the surement of digital radiographs,18 multidetector computed to-
measurement of all skeletal elements that contribute to living mography images,19 however, mostly are recorded from direct
height.1,2 Although highly accurate, this anatomical method cannot measurement of cadavers.20e24 The rapid technological develop-
be applied in many cases in which human remains have been ment of multi-row computed tomography (CT) in the last two de-
dismembered or mutilated. As a result, the most commonly alter- cades has resulted in an increase in CT examinations on the whole,
natives are based on the positive linear relationship between stature the volume rendering technique (VRT) has the advantage of real
and the length of various parts of the body. Several post-cranial time operation of the object mass. Currently, the virtual anthro-
pology techniques initiated by the Virtopsy (virtual autopsy) proj-
ect,25 which rely on three-dimensional (3D) reconstructions of
computed tomography (CT) scans of skeletal remains. Analysts can
* Corresponding author. Tel.: þ86 2885501553; fax: þ86 2885501544.
perform morphological and metric analyses to establish the bio-
E-mail address: [email protected] (Z.-h. Deng).
g
These authors contributed equally to this work.
logical profile from the 3D images,26 and the data can provide an

http://dx.doi.org/10.1016/j.jflm.2015.07.003
1752-928X/© 2015 Elsevier Ltd and Faculty of Forensic and Legal Medicine. All rights reserved.
 K. Zhang et al. / Journal of Forensic and Legal Medicine 35 (2015) 40e44 41

objective and observer-independent method for recording features

that are normally difficult to quantify and can allow easy access to
reports and data for peer review. Additionally, advantages include
no deterioration of data over time, no skeletal preparation, no
damage to skeletal material, and potential application to living
individuals.27,28
Therefore, the primary objective of the present investigation
was to generate linear regression models for stature estimation on
the basis of sternum length derived from 3D-VRT images in
different population groups. This was done due to inherent popu-
lation differences in various dimensions that are attributed to ge-
netic and environmental factors.29,30 Furthermore, a test of the
utility of sternal measurements to predict stature in Western Chi-
nese was carried out.

2. Materials and methods

The present study was performed with the approval of the

ethics committee of the West China Hospital of Sichuan University
and all the participants provided written informed consent.
The data used in the present study were obtained from partic-
ipants undergoing routine examination at the West China Hospital
of Sichuan University. The study sample comprised 288 individuals
from western China, including 124 females and 164 males, with
documented ages between 19 and 78 years (Table 1). Subjects with
a history of chronic illness, trauma, physical deformity, or any
surgical procedure that might affect stature or sternum dimensions
were excluded from the study.
The stature was measured during standing in full extension as
the maximum length between the skull vertex and the heel in
millimetres. The data collection was conducted during a two-
month period between 9:00 am and 10:00 am to avoid the influ-
ence of diurnal variation as it affects the standards generated and
equations developed for the estimation of stature.31 Fig. 1. Linear measurements (black lines) from 3D-VRT images used to calculate the
Multidetector CT was carried out on Somatom definition AS 128 combined length of manubrium and mesosternum: manubrium length (M) and
mesosternum length (B).
slice CT Machine by Siemens Germany Ltd. After obtaining the
scout projection, the area of scanning was defined to include the
region from fifth cervical vertebra until the first lumbar vertebra. later life,32 and as a result, the Combined length of manubrium and
The scanning protocol was as follows: collimation of 1 mm, mesosternum was used for statistical analysis.
reconstruction interval of 1 mm, tube voltage of 120 kV, tube cur- All the measurements recorded by the same author twice to
rent of 110 mA, and scanning time of 0.3 s. Image data were pro- minimize the error in measurement, and the results of measure-
cessed on a workstation (Syngo CT 2011A) to obtain volume- ments were made to the nearest 1 mm, the mean value was used
rendering technique images. for analysis. Mean intra-observer error, calculated following the
A 3D reconstructed image was used for assessment. The linear equation presented by Albanese et al.,33,34 was less than 1.1% for all
distances from the centre of suprasternal notch or incisura jugularis dimensions and thus within the acceptable limit (below 2.0e2.5%).
(jugular notch) to the centre of manubrio-mesosternal junction Given that previous studies have demonstrated that stature
(sternal synchondrosis) in mid-sagittal plane and the manu- formulae are not only population-specific but also sex-spe-
brioemesosternal junction to the mesosternoexiphoidal junction cific,23,35,36 separate linear regression equations for estimating
of the sternum in the mid-sagittal plane were defined as the Length stature on the basis of sternum length were then devised for males
of manubrium (M) and Length of mesosternum (B), respectively and females. The study sample was randomly divided into two
(Fig. 1). The manubrium and mesosternum often do not fuse even in subgroups. A calibration sample, which consisted of 113 females

Table 1
Sex, age, stature, and combined length of manubrium and mesosternum distribution of the utilized study sample.

Sex group n Age (years) Stature (cm) Combined length of manubrium and
mesosternum (cm)

Mean ± Std. deviation Minimum Maximum Mean ± Std. deviation Minimum Maximum Mean ± Std. deviation Minimum Maximum

Calibration sample

Female 113 47.0 ± 12.3 19 78 157.1 ± 6.5 138.9 171.9 12.9 ± 1.0 10.0 15.8
Male 146 48.9 ± 12.1 25 78 166.6 ± 6.3 151.9 185.9 14.8 ± 1.4 9.5 23.0

Validation sample

Female 11 45.5 ± 8.3 33 59 156.5 ± 3.2 150.0 160.4 12.6 ± 0.9 11.1 14.4
Male 18 47.9 ± 7.7 35 60 168.1 ± 2.5 163.5 171.4 14.8 ± 0.9 13.2 16.3
42 K. Zhang et al. / Journal of Forensic and Legal Medicine 35 (2015) 40e44

and 146 males, was used to develop the sex-specific linear individuals fell within the 95% confidence interval of the predicted
regression equations. A validation sample comprising the remain- value (Table 2).
ing 11 female and 18 male individuals was then used to test the
predictive accuracy of the two models. The normality of the sternal
4. Discussion
length data for both sex samples was verified with KeS test
(Z ¼ 0.851, P ¼ 0.464 for female, and Z ¼ 0.893, P ¼ 0.402 for male).
Recent studies have evaluated the potential of using the 3D CT
For assessing the correlation between the stature and combined
images to estimate stature from pelvic bones,37 radial and ulnar
length of manubrium and mesosternum, the Pearson's correlation
lengths,38 and scapula,9 examining 3D reconstructions of pelvic CT
coefficient was calculated and its significance was tested by t test. P
scans of live adults to estimate age,39 and applying of postmortem
value of less than 0.05 was considered as significant. Specifically,
3D-CT facial reconstruction for personal identification.16 The pre-
model accuracy was assessed by observing the number of in-
sent investigation demonstrate that combined length of the ma-
dividuals in the validation sample whose corrected stature fell
nubrium and mesosternum derived from 3D-VRT has a statistically
within the 95% confidence interval of the predicted value, following
significant association with stature in the Western Chinese. To date,
the developed equation. The data were analysed by using the SPSS
some researchers have applied CT scanning to the field of anthro-
(VERSION 13.0 for Windows).
pology, revealing that there was no significant difference between
the actual bone dimensions and the measurements on the 3D CT
3. Results images.37,40 In addition, CT enables forensic investigators to
perform measurements on bone images instead of fresh bones that
Descriptive statistics for the calibration sample and validation require preparations such as removing and cleansing the skeleton if
sample was shown in Table 1. surrounding tissues remain.17 Moreover, digital CT images can be
The linear regression analysis for the calibration sample data preserved almost permanently in little or no physical space,
yielded the following formulae: male stature (cm) ¼ 137.28 þ enabling repeated measurements and potentially decreasing
1.99*combined length of manubrium and mesosternum and female intraobserver and interobserver error.26
stature (cm) ¼ 111.59 þ 3.51*combined length of manubrium and It is well known that cadaver stature is different from living
mesosternum. Combined length of manubrium and mesosternum stature41 because of rigor mortis and positional differences.23,42 A
showed a significant positive correlation with stature in both males number of previous studies3,5,43e45 have demonstrated that an
and females (p < 0.01). Pearson's correlation coefficients for the adjustment factor should be applied to cadaver length given that
regression models were r ¼ 0.459 and r ¼ 0.541 for the male and living height is consistently less than the value recorded postmor-
female formulae, respectively. The standard errors of the estimate tem. This discrepancy is primarily due to the compression of inter
were 4.76 cm for the male equation and 6.73 cm for the female vertebral disks and other joints when a person is standing, the
equation. position adopted for the measurement of stature in the living.43,44 It
When the combined length of manubrium and mesosternum is generally accepted that the extent to which the body lengthens
data of the validation sample were applied to the above derived after death is approximately 2.0 cm.3,5,45 The regression model in
linear regression equations, the corrected stature of all 29 the present study was based on living standing height.

Table 2
Accuracy assessment of the derived stature equations when applied to the validation sample.

Sex Age (years) Combined length of manubrium and mesosternum (cm)

Corrected stature Predicted stature Lower 95% CI Upper 95% CI

Female 59 150.1 153.2 139.7 166.6

Female 52 157.5 155.5 142.0 168.9
Female 38 160.5 156.3 142.8 169.7
Female 43 159.5 156.3 142.8 169.7
Female 43 153.5 153.0 139.5 166.4
Female 34 158.4 158.1 144.6 171.5
Female 33 158.5 158.2 144.7 171.6
Female 51 152.3 150.4 136.9 163.8
Female 44 155.8 152.8 139.3 166.2
Female 51 158.2 162.2 148.7 175.6
Female 52 157.2 156.7 143.2 170.1
Male 51 171.4 157.6 162.4 176.6
Male 48 169.3 155.7 160.4 174.7
Male 60 167.5 156.3 161.0 175.3
Male 59 169.9 158.0 162.8 177.0
Male 50 168.8 157.8 162.5 176.8
Male 51 167.0 157.0 161.8 176.0
Male 56 170.1 160.3 165.0 179.3
Male 39 163.5 154.4 159.2 173.4
Male 35 168.0 157.2 161.9 176.2
Male 45 170.2 159.6 164.3 178.6
Male 43 165.2 156.8 161.6 175.8
Male 42 170.2 158.7 163.5 177.7
Male 53 171.0 160.1 164.8 179.1
Male 43 165.7 156.2 160.9 175.2
Male 44 171.2 157.9 162.6 176.9
Male 51 165.1 155.9 160.6 174.9
Male 35 167.5 153.9 158.7 172.9
Male 57 164.2 156.6 161.3 175.6
 K. Zhang et al. / Journal of Forensic and Legal Medicine 35 (2015) 40e44 43

The estimate for the stature formulae derived from the sternum 10. Giurazza F, Del Vescovo R, Schena E, Cazzato RL, D'Agostino F, Grasso RF, et al.
Stature estimation from scapular measurements by CT scan evaluation in an
length in the current study are generally comparable to those
Italian population. Leg Med Tokyo 2013;15(4):202e8.
previous investigations in other population groups. For example, 11. Campobasso CP, Di Vella G, Introna Jr F. Using scapular measurements in
Menezes et al. reported a standard error of 5.64 cm for South Indian regression formulae for the estimation of stature. Boll Soc Ital Biol Sper
males24 and 4.11 cm for South Indian females.20 Similarly, Singh 1998;74(7e8):75e82.
12. Sagir M. Estimation stature from X-rays of metacarpals in the Turkish popu-
et al.23 in their study of a Northwest Indian population sample, lation. Anthropol Anz 2006;64(4):377e88.
obtained a standard error of 6.83 cm for males and 6.65 cm for 13. Bidmos MA. Metatarsals in the estimation of stature in South Africans.
females, respectively. Macaluso and Lucena18 reported the standard J Forensic Leg Med 2008;15(8):505e9.
14. Nagesh KR, Pradeep Kumar G. Estimation of stature from vertebral column
errors of 6.20 cm for the males and 5.56 cm for the females in length in South Indians. Leg Med Tokyo 2006;8(5):269e72.
Spanish population and the correlation coefficients for the regres- 15. Terazawa K, Takatori T, Mizukami K, Tomii S. Estimation of stature from
sion models were r ¼ 0.391 and r ¼ 0.461 for the male and female somatometry of vertebral column in Japanese. Nihon Hoigaku Zasshi
1985;39(1):35e40.
formulae, respectively. Yonguc et al.22 derived regression equations 16. Tibbetts GL. Estimation of stature from the vertebral column in American
for estimating stature from measured sternal lengths in Turks and Blacks. J Forensic Sci 1981;26(4):715e23.
the length of the manubrium and body had the highest correlation 17. Torimitsu S, Makino Y, Saitoh H, Ishii N, Hayakawa M, Yajima D, et al. Stature
estimation in Japanese cadavers using the sacral and coccygeal length
coefficient in both males and females (correlation coefficient: 0.721 measured with multidetector computed tomography. Leg Med Tokyo
and 0.740, respectively). The present study concludes that the 2014;16(1):14e9.
length of the sternum is a reliable predictor of stature in Chinese 18. Macaluso Jr PJ, Lucena J. Stature estimation from radiographic sternum length
in a contemporary Spanish population. Int J Leg Med 2014;128(5):845e51.
and can be used as a tool for stature estimation when better pre-
19. Torimitsu S, Makino Y, Saitoh H, Sakuma A, Ishii N, Hayakawa M, et al. Stature
dictors of stature like the long bones are not available when estimation based on measurements of the sternal medullary cavity using
examining skeletal remains. multidetector computed tomography images of Japanese cadavers. Forensic Sci
Int 2014;242:299. e291e295.
20. Menezes RG, Nagesh KR, Monteiro FN, Kumar GP, Kanchan T, Uysal S, et al.
5. Conclusions Estimation of stature from the length of the sternum in South Indian females.
J Forensic Leg Med 2011;18(6):242e5.
21. Marinho L, Almeida D, Santos A, Cardoso HF. Is the length of the sternum
The regression equations derived from the sternum length in
reliable for estimating adult stature? A pilot study using fresh sterna and a test
the present study may be used for stature estimation in Western of two methods using dry sterna. Forensic Sci Int 2012;220(1e3):292.
Chinese although they can only provide a wide 95% confidence e291e294.
intervals, and the CTeVRT method can be a practical method for 22. Yonguc GN, Kurtulus A, Bayazit O, Adiguzel E, Unal I, Demir S, et al. Estimation
of stature and sex from sternal lengths: an autopsy study. Anat Sci Int
stature estimation. 2015;90(2):89e96.
23. Singh J, Pathak RK, Chavali KH. Skeletal height estimation from regression
analysis of sternal lengths in a Northwest Indian population of Chandigarh
Conflict of interest
region: a postmortem study. Forensic Sci Int 2011;206(1e3):211. e211e218.
None to declare. 24. Menezes RG, Kanchan T, Kumar GP, Rao PP, Lobo SW, Uysal S, et al. Stature
estimation from the length of the sternum in South Indian males: a preliminary
study. J Forensic Leg Med 2009;16(8):441e3.
Funding
25. Thali MJ, Yen K, Schweitzer W, Vock P, Boesch C, Ozdoba C, et al. Virtopsy, a
None declared. new imaging horizon in forensic pathology: virtual autopsy by postmortem
multislice computed tomography (MSCT) and magnetic resonance imaging
(MRI) e a feasibility study. J Forensic Sci 2003;48(2):386e403.
Ethical approval 26. Grabherr S, Cooper C, Ulrich-Bochsler S, Uldin T, Ross S, Oesterhelweg L, et al.
None. Estimation of sex and age of “virtual skeletons” e a feasibility study. Eur Radiol
2009;19(2):419e29.
27. Telmon N, Gaston A, Chemla P, Blanc A, Joffre F, Rouge D. Application of the
Acknowledgement Suchey-Brooks method to three-dimensional imaging of the pubic symphysis.
J Forensic Sci 2005;50(3):507e12.
This work was supported by the National Natural Science 28. Sidler M, Jackowski C, Dirnhofer R, Vock P, Thali M. Use of multislice computed
tomography in disaster victim identification e advantages and limitations.
Foundation of China (No. 81373252) and the Applied Basic Research Forensic Sci Int 2007;169(2e3):118e28.
Programs of Science and Technology Commission Foundation of 29. Kanchan T, Menezes RG, Kotian MS. Stature estimation: valuable precautions.
Sichuan Province (No. 2013JY0148). J Forensic Leg Med 2008;15(6):413.
30. Krishan K, Sharma A. Estimation of stature from dimensions of hands and feet
in a North Indian population. J Forensic Leg Med 2007;14(6):327e32.
References 31. Krishan K, Kanchan T, Menezes RG, Ghosh A. Forensic anthropology casework-
essential methodological considerations in stature estimation. J Forensic Nurs
1. Raxter MH, Auerbach BM, Ruff CB. Revision of the fully technique for esti- 2012;8(1):45e50.
mating statures. Am J Phys Anthropol 2006;130(3):374e84. 32. Chandrakanth HV, Kanchan T, Krishan K, Arun M, Pramod Kumar GN. Esti-
2. Raxter MH, Ruff CB, Auerbach BM. Technical note: revised fully stature esti- mation of age from human sternum: an autopsy study on a sample from South
mation technique. Am J Phys Anthropol 2007;133(2):817e8. India. Int J Leg Med 2012;126(6):863e8.
3. Mahakkanukrauh P, Khanpetch P, Prasitwattanseree S, Vichairat K, Troy Case D. 33. Albanese J, Eklics G, Tuck A. A metric method for sex determination using the
Stature estimation from long bone lengths in a Thai population. Forensic Sci Int proximal femur and fragmentary hipbone. J Forensic Sci 2008;53(6):1283e8.
2011;210(1e3):279. e271e277. 34. Albanese J. A metric method for sex determination using the hipbone and the
4. Smith SL. Stature estimation of 3-10-year-old children from long bone lengths. femur. J Forensic Sci 2003;48(2):263e73.
J Forensic Sci 2007;52(3):538e46. 35. Bidmos M, Asala S. Calcaneal measurement in estimation of stature of South
5. Radoinova D, Tenekedjiev K, Yordanov Y. Stature estimation from long bone African blacks. Am J Phys Anthropol 2005;126(3):335e42.
lengths in Bulgarians. Homo 2002;52(3):221e32. 36. Sahni D, Sanjeev, Sharma P, Harjeet, Kaur G, Aggarwal A. Estimation of stature
6. Munoz JI, Linares-Iglesias M, Suarez-Penaranda JM, Mayo M, Miguens X, from facial measurements in northwest Indians. Leg Med Tokyo 2010;12(1):23e7.
Rodriguez-Calvo MS, et al. Stature estimation from radiographically deter- 37. Torimitsu S, Makino Y, Saitoh H, Sakuma A, Ishii N, Hayakawa M, et al. Stature
mined long bone length in a Spanish population sample. J Forensic Sci estimation in Japanese cadavers based on pelvic measurements in three-
2001;46(2):363e6. dimensional multidetector computed tomographic images. Int J Legal Med
7. Ahmed AA. Estimation of stature using lower limb measurements in Sudanese 2015;129(3):633e9.
Arabs. J Forensic Leg Med 2013;20(5):483e8. 38. Torimitsu S, Makino Y, Saitoh H, Sakuma A, Ishii N, Hayakawa M, et al. Stature
8. Nor FM, Abdullah N, Mustapa AM, Qi Wen L, Faisal NA, Ahmad Nazari DA. estimation based on radial and ulnar lengths using three-dimensional images
Estimation of stature by using lower limb dimensions in the Malaysian pop- from multidetector computed tomography in a Japanese population. Leg Med
ulation. J Forensic Leg Med 2013;20(8):947e52. Tokyo 2014;16(4):181e6.
9. Torimitsu S, Makino Y, Saitoh H, Sakuma A, Ishii N, Hayakawa M, et al. Stature 39. Wink AE. Pubic symphyseal age estimation from three-dimensional re-
estimation in Japanese cadavers based on scapular measurements using mul- constructions of pelvic CT scans of live individuals. J Forensic Sci 2014;59(3):
tidetector computed tomography. Int J Leg Med 2015;129(1):211e8. 696e702.
44 K. Zhang et al. / Journal of Forensic and Legal Medicine 35 (2015) 40e44

40. Sakuma A, Ishii M, Yamamoto S, Shimofusa R, Kobayashi K, Motani H, et al. distinguishing birth cohort effects from actual loss of stature with aging. Hum
Application of postmortem 3D-CT facial reconstruction for personal identifi- Biol 1989;61(3):415e25.
cation. J Forensic Sci 2010;55(6):1624e9. 43. Trotter M, Gleser GC. Estimation of stature from long bones of American
41. Giurazza F, Del Vescovo R, Schena E, Battisti S, Cazzato RL, Grasso FR, et al. Whites and Negroes. Am J Phys Anthropol 1952;10(4):463e514.
Determination of stature from skeletal and skull measurements by CT scan 44. Hauser R, Smolinski J, Gos T. The estimation of stature on the basis of mea-
evaluation. Forensic Sci Int 2012;222(1e3):398. e391e399. surements of the femur. Forensic Sci Int 2005;147(2e3):185e90.
42. Cline MG, Meredith KE, Boyer JT, Burrows B. Decline of height with age in 45. De Mendonca MC. Estimation of height from the length of long bones in a
adults in a general population sample: estimating maximum height and Portuguese adult population. Am J Phys Anthropol 2000;112(1):39e48.