Stature estimation from the scapula measurements using 3D-volume rendering technique by regression equations in the Northern Indian population

Abstract

The present study assessed the correlation between the stature and scapular measurements from both sides in order to develop population-specific regression equations to estimate the stature from measurements of the scapula in a contemporary Northern Indian population individually for the left and right sides. A total of 597 cadavers underwent postmortem multidetector computed tomography and subsequent medicolegal autopsy in our department between August 2021 and August 2022. Two hundred samples (100 males and 100 females) were randomly collected based on inclusion and exclusion criteria. Six linear anthropometric measurements of the scapula from either side were measured using the 3D volume-rendered technique by an electronic cursor. Each anthropometric measurement showed a significant difference between males and females (p < 0.05). Regression analysis was applied to match the taken measurements against stature. The accuracy to predict stature ranged from 3.99 to 4.94 cm for males and from 4.49 to 5.27 cm for females, respectively. Left-sided measurements were better predictors of stature than the right side in both genders. The results of this study indicate that scapular measurements could be useful to estimate the stature of Northern Indian individuals, particularly in scenarios of disaster victim identification lacking long bones, which are considered to be better predictors to date.

Keywords

Disaster victim identification forensic science forensic anthropology scapula measurements stature estimation 3D volume-rendered technique

Introduction

Stature is considered a vital parameter among the ‘big four’ (age, sex, stature, and race) in cases of disaster victim identification. The estimation of stature from the remains is a challenging aspect to any forensic anthropologist/pathologist because the opinion given guides the investigating authorities. Although the exact stature of the particular bone is difficult to be measured, a specific stature range could be given based on anthropometric measurements since they are biologically related. But the universality of the application of the derived formulae from measurements of a particular population to a different population is still unclear. This mandates the researchers to explore population-specific and gender-specific formulae. Globally, the literature for stature estimation from the scapula is less available because it is placed under several layers of back muscles resulting in less exploration by the researchers. The invention of postmortem multidetector computed tomography (PMCT) has made it possible to study any part of the body with intact anatomy of the deceased. Studies are conducted in various bones of the body from head to toe and corresponding regression equations are derived for the respective populations using PMCT.^1–6 The facts that after the completion of skeletal development during life; morphological changes of the scapula are negligible⁷ and smaller flat bones are better preserved in any type of disaster had drawn the attention to derive a population-specific formula to estimate the stature from the scapula.⁸ Hence, the authors aimed to provide a separate metric and statistical analysis from both sides of scapula measurements in the Northern India population living in Delhi.

Methods

Sample collection

This cross-sectional study was conducted under the project of the Centre for Advanced Research and Excellence in Virtual Autopsy commenced in the middle of August 2021 at the Department of Forensic Medicine and Toxicology, All India Institute of Medical Sciences [AIIMS], New Delhi. This is a collaboration established by the two-premier institutes of the country, that is, the Indian Council of Medical Research, New Delhi, and AIIMS, New Delhi. The data were collected from the deceased subjects who underwent PMCT examination as a part of routine autopsy examination. There were cases such as road traffic accidents, falls from height, gunshots, electrocution, railway accidents, poisoning, hanging, strangulation, and sudden death cases. Northern India has the following states: Delhi, Haryana, Himachal Pradesh, Jammu and Kashmir, Uttar Pradesh, and Uttarakhand based on the place of their birth and their parents’ birth; however, the ancestral origin was not known in most of the cases. The study sample comprised the age group from 21 years of age to 78 years of age who underwent PMCT examination as a part of an autopsy examination was included in the study. The study participants who had sustained a fracture of any bones of the body modifying the stature, with any deformity of the spine altering the stature, with notable gross congenital scapula defects, or sustained scapula fracture altering the measurements were excluded from the study. A total of 200 deceased between September 2021 and September 2022 were collected from the people belonging to mentioned regions of North India who had succumbed to death due to any of the causes abiding by the inclusion and exclusion criteria. A temporary identification was given along with details like postmortem no, age, sex, and locality, and the same was entered initially in a data collection proforma.

Anthropometric measurements

Postmortem multidetector computed tomography examination was performed using a 16-slice MSCT spiral scanner, Cannon America Medical Systems, Inc Aquilion Lightning TSX-035A CT. Scanning parameters were 120 kV and 70 mAs; 16 × 1 mm collimation was used for all the cases for data acquisition. All the raw data were processed into slices of 1 mm thickness. The results of the study were evaluated with the Vitrea software v.6.9.1. The images were viewed using the 3D VRT. A total of 12 variables from both sides (six from each side) were measured from the reconstructed images as shown in Figure 1(a) to (c). The left-sided and right-sided measurements are abbreviated as names of the measurements following L and R, respectively. All the variables were measured using the electronic cursor [distance tool] available in the software by a forensic medicine specialist who had acquired training and experience in radiology. The collected data were entered directly into a data collection proforma initially followed which was updated in Microsoft Excel 2016 Spreadsheet.

Figure 1.

(a) 3D-volume rendering technique (VRT) to measure stature using postmortem multidetector computed tomography (PMCT). (b) and (c) 3D VRT to measure the different variables of the scapula. Scapula Morphological Width (SMW), Scapula Morphological Length (SML), Longitudinal Scapula Length (LSL), Longitudinal Maximum Length (LML), Transverse Scapula Length (TSL), and Axillary Margin Length (AML). (d) Gender-wise distribution of the total no of participants.

STAT

The starting point on the head end was from the most distal part of the vertex and the ending part in the lower limb was the distal-most part of the calcaneum.

Scapula measurements

The following measurements are taken from both the left and right scapula by volume-rendered technique (VRT) using the electronic cursor.

Scapula Morphological Width (SMW): It is measured as the distance between the upper-end medial margin and the middle part of the glenoid cavity.

Scapula Morphological Length (SML): It is measured as the distance between the tip of the inferior angle and the tip of the superior angle of the scapula.

Longitudinal Scapula Length (LSL): It is measured as the distance between the tip of the inferior angle and the tip of the superior margin of the coracoid process.

Longitudinal Maximum Length (LML): It is measured as the distance between the tip of the inferior angle and the tip of the superior margin of the acromion process.

Transverse Scapula Length (TSL): It is measured as the distance between the upper medial margin and the tip of the inferior margin of the glenoid cavity.

Axillary Margin Length (AML): It is measured as the distance between the tip of the inferior angle and the tip of the inferior margin of the glenoid cavity.

Statistical analysis

The intra-observer reliability was assessed by using the test–retest method by measuring the length 2 times by the same observer at different periods of time. Inter-observer reliability was carried out by comparing the average obtained by the first observer with measurements taken by the second independent observer for 50 cases. All lengths were measured in centimeters to the nearest 0.1 mm. The obtained data were tabulated and analyzed intraclass correlation (ICC) method. The descriptive data of scapular measurements namely SMW, SML, LSL, LML, TSL, and AML from both sides and STAT obtained using the PMCT tool were expressed as mean with SD and range. The linear relationship between the STAT and individual measurements had been explored by using correlation analysis. Regression analysis both linear and multiple was used to establish the relationship between STAT and individual parameters. The derived equations were validated in the studied sample populations using reliability analysis. All statistical analyses were carried out at a 5% level of significance using IBM Statistical Package for Social Sciences (SPSS) (v.23.0).

Results

Descriptive statistics

Two hundred subjects participated in this study from 2021 to 2022 comprising 100 males and females each. The randomly collected subjects were segregated according to age and gender as presented in Figure 1(d). The descriptive statistics for the general population, males and females are presented in Table 1. The independent variable age was mentioned in years, all the other outcome variables were measured in cm. The mean age of the female population was higher than the mean age of the male population that participated in the study. The mean difference in stature between males and females was 13.7 cm with a male preponderance. A significant difference in the lengths of both side scapula measurements was observed between males and females (p < 0.05). The right-sided mean length of the scapula measurements was relatively higher than the left side in both genders.

Table 1.

Descriptive data in relation to different study characteristics – combined and gender-wise.

	Combined		Male		Female
Sex/Parameters	Mean (SD)	Range	Mean (SD)	Range	Mean (SD)	Range
Age	40.46 (15.62)	21–78	39.49 (15.2)	21–78	41.43 (16.05)	21–78
Stat_CT	159.05 (8.96)	138.5–180.9	165.7 (6.01)	154.6–180.9	152.5 (6.1)	138.5–166.1
LSMW	10.06 (1.24)	7.64–14.19	10.8 (1.18)	8.77–14.91	9.33 (0.77)	7.64–10.95
LSML	14.16 (1.54)	10.16–17.69	15.28 (1.1)	12.12–17.69	13.06 (1.02)	10.16–15.59
LLSL	15.91 (1.35)	12.46–18.85	16.81 (0.99)	14.03–18.85	15.02 (1.04)	12.46–17.93
LTSL	10.41 (1.05)	7.72–13.68	11.02 (0.92)	9.23–13.68	9.8 (0.77)	7.72–11.75
LLML	17.67 (1.33)	14.57–20.79	18.55 (1.01)	16.07–20.79	16.8 (0.98)	14.57–18.9
LAML	12.78 (1.16)	9.62–15.65	13.28 (0.87)	11.23–14.84	12.28 (1.2)	9.62–15.65
RSMW	10.14 (1.19)	7.44–14.75	10.86 (1.14)	8.59–14.75	9.41 (0.7)	7.44–11.2
RSML	14.25 (1.62)	10.02–17.79	15.44 (1.17)	13.04–17.79	13.06 (1.01)	10.02–15.4
RLSL	15.95 ± 1.34	12.39–19.98	16.86 (0.96)	14.88–19.98	15.04 (1)	12.39–17.06
RTSL	10.43 ± 1.01	7.64–13.32	11.0 (0.88)	9.36–13.32	9.85 (0.79)	7.64–11.91
RLML	17.72 ± 1.32	14.53–20.68	18.83 (0.96)	16.13–20.68	16.8 (0.93)	14.53–18.75
RAML	12.73 ± 1.16	10.15–215.85	13.23 (0.89)	11.03–14.93	12.22 (1.19)	10.15–15.85

* Age in years, all measurements are in cms.

Intra-observer error and inter-observer

The intra-observer reliability was assessed by using the test-retest method by measuring the length two times by the same observer at different periods of time. The second measurement was taken after a span of seven days post-first measurement and the mean value of the two was considered for analysis. Inter-observer reliability was carried out by comparing the average obtained by the first observer with measurements taken by the second independent observer for 50 cases. The mean value calculated by the second observer was considered for the reliability analysis. The observers’ reliability was explored using ICC coefficient method. It was observed more than 96% of consistency is noted between the observers for all the measurements (Table 2).

Table 2.

Intraobserver and interobserver analysis by using intraclass correlation method (ICC).

Measurement	ICC – Intra-observer Analysis	ICC – Inter-observer Analysis
LSMW	0.970	0.975
LSML	0.989	0.978
LLSL	0.975	0.970
LTSL	0.985	0.989
LLML	0.990	0.985
LAML	0.966	0.968
RSMW	0.968	0.973
RSML	0.983	0.976
RLSL	0.982	0.975
RTSL	0.981	0.990
RLML	0.986	0.988
RAML	0.963	0.970

Regression analysis

Table 3 shows the regression equations calculated for each variable in relation to stature for the combined population and each sex. All measurements presented statistically significant correlation coefficients with stature (p < 0.05). The accuracy of stature prediction ranged from 3.99 to 4.94 cm for males, and 4.49 to 5.27 cm for female when stature was calculated using the equations that involved the dimensions of a single variable. The right longitudinal maximum length gave the most accurate measurement considering the combined population, while the gender-wise comparison showed the left-sided measurements were more accurate than the right-sided measurements. The left scapula morphological length gave the most accurate results for males and the left longitudinal maximum length gave the most accurate results for females. Multiple regression equations are presented in Table 4, with the accuracy of stature prediction being more for the left-sided scapula than the right-sided scapula among the combined population, male and female populations. The range of accuracy for stature prediction using multiple regression equations was from 3.44 to 3.90.

Table 3.

Linear regression equations for estimation of stature from various scapular measurements among combined population (p < 0.001), males (p < 0.001) and males (p < 0.001) on both left and right sides.

Equation	R	R ²	SEE	95% CI	ICC
Combined (n = 200)
117.44 + 4.07* (LSMW)	0.61	0.38	7.09	3.34–4.81	0.75
98.39 + 4.24 * (LSML)	0.75	0.56	5.97	3.71–4.77	0.89
82.66 + 4.76* (LLSL)	0.72	0.52	6.21	4.12–5.40	0.82
68.62 + 5.08* (LLML)	0.76	0.58	5.82	4.47–5.69	0.87
99.91 + 5.62 * (LTSL)	0.68	0.47	6.56	4.78–6.47	0.80
90.79 + 5.32* (LAML)	0.67	0.45	6.68	4.49–6.15	0.75
114.74 + 4.31* (RSMW)	0.62	0.39	7.04	3.55–5.08	0.87
101.32 + 4.01 * (RSML)	0.74	0.55	6.06	3.5–4.5	0.86
83.75 + 4.68* (RLSL)	0.71	0.51	6.31	4.03–5.32	0.80
66.39 + 5.19* (RLML)	0.77	0.59	5.74	5.11–6.87	0.86
96.10 + 5.99 * (RTSL)	0.69	0.48	6.49	4.59–5.8	0.78
93.49 + 5.12 * (RAML)	0.65	0.42	6.83	4.28–5.97	0.71
Males (n = 100)
129.35 + 3.36* (LSMW)	0.66	0.44	4.54	2.59–4.13	0.75
102.91 + 4.11 * (LSML)	0.75	0.56	3.99	3.38–4.83	0.89
101.07 + 3.84 * (LLSL)	0.63	0.40	4.70	2.89–4.79	0.80
86.14 + 4.28* (LLML)	0.72	0.52	4.17	3.46–5.11	0.86
115.41 + 4.56 * (LTSL)	0.70	0.49	4.33	3.62–5.5	0.79
108.64 + 4.29 * (LAML)	0.62	0.39	4.74	3.20–5.38	0.72
129.6 + 3.32* (RSMW)	0.63	0.40	4.70	2.5–4.14	0.85
109 + 3.67 * (RSML)	0.72	0.51	4.23	2.95–4.39	0.89
104.42 + 3.63 * (RLSL)	0.58	0.36	4.93	2.61–4.66	0.76
86.38 + 4.25* (RLML)	0.68	0.46	4.43	3.34–5.17	0.85
115.57 + 4.55 * (RTSL)	0.67	0.44	4.51	3.53–5.57	0.77
114.86 + 3.88 * (RAML)	0.58	0.33	4.94	2.78–4.98	0.70
Females (n = 100)
107.52 + 4.82 * (LSMW)	0.61	0.37	4.87	3.56–6.08	0.81
106.99 + 3.48 * (LSML)	0.58	0.34	4.98	2.51–4.45	0.88
97.11 + 3.69 * (LLSL)	0.63	0.39	4.78	2.77–4.61	0.84
81.45 + 4.23 * (LLML)	0.68	0.46	4.49	3.32–5.14	0.87
100.52 + 5.30 * (LTSL)	0.67	0.45	4.56	4.12–6.48	0.84
110.39 + 3.43 * (LAML)	0.67	0.45	4.54	2.67–4.18	0.81
102.06 + 5.35 * (RSMW)	0.62	0.38	4.83	3.98–6.73	0.87
112.05 + 3.09 * (RSML)	0.51	0.26	5.27	2.06–4.13	0.85
96.50 + 3.72 * (RLSL)	0.61	0.37	4.86	2.75–4.69	0.81
78.68 + 4.39 * (RLML)	0.67	0.45	4.57	3.42–5.37	0.87
100.85 + 5.24 * (RTSL)	0.68	0.46	4.53	4.09–6.39	0.83
111.04 + 3.39 * (RAML)	0.66	0.44	4.60	2.62–4.16	0.79

SEE: standard error estimate; R: correlation coefficient; ICC: intraclass correlation.

Table 4.

Multiple regression equations for estimation of stature from various scapular measurements among combined population, males and females on both left and right side.

Side	R	R ²	SEE	MRE	ICC
Combined (n = 200)
Both side	0.91	0.83	3.82	63.85 + 0.07 (LSMW) + 1.5 (LSML) – 1 (LLSL) + 1.49 (LLML) + 1.31 (LTSL) + 0.65 (LAML) + 1.03 (RSMW) + 0.6 (RSML) + 0.8 (RLSL) + 0.4(RLML) + 0.27(RTSL)–0.04(RAML)	0.95
Left	0.91	0.82	3.82	64.86 + 0.8 (LSMW) + 2.19 (LSML) – 0.35 (LLSL) + 1.98 (LLML) + 1.68 (LTSL) + 0.64 (LAML)	0.95
Right	0.90	0.82	3.90	64.14 + 1.52 (RSMW) + 1.67 (RSML) + 0.4 (RLSL) + 1.73 (RLML) + 1.15 (RTSL) + 0.53 (RAML)	0.95
Male Population (n = 100)
Both side	0.84	0.70	3.50	88.36 + 1.23 (LSMW) + 1.29 (LSML)–0.89 (LLSL) + 2.13 (LLML) + 1.40 (LTSL) + 0.77 (LAML)–0.34 (RSMW) + 0.92 (RSML) + 0.18(RLSL)–0.92 (RLML)–0.39(RTSL) + 0.13(RAML)	0.90
Left	0.83	0.69	3.44	86.63 + 0.94 (LSMW) + 1.9 (LSML) – 0.92 (LLSL) + 1.68 (LLML) + 1.17 (LTSL) + 0.83 (LAML)	0.88
Right	0.81	0.65	3.65	85.81 + 1.26 (RSMW) + 1.65 (RSML) + 0.26 (RLSL) + 0.85 (RLML) + 0.82 (RTSL) + 0.87(RAML)	0.90
Female Population (n = 100)
Both side	0.81	0.65	3.83	71.85–1.09 (LSMW) + 2.66 (LSML)–1.04 (LLSL) + 1.59 (LLML) + 1.09 (LTSL)–0.12 (LAML) + 1.67 (RSMW) – 1.158 (RSML) + 0.25(RLSL) + 0.21 (RLML) + 1.27(RTSL) + 1.25(RAML)	0.88
Left	0.80	0.63	3.82	73.88 + 0.23 (LSMW) + 1.5 (LSML) – 0.65 (LLSL) + 1.83 (LLML) + 2.34 (LTSL) + 1.06 (LAML)	0.87
Right	0.79	0.62	3.87	70.22 + 1.36 (RSMW) + 1.1 (RSML) – 0.46 (RLSL) + 1.88 (RLML) + 1.9 (RTSL) + 0.95 (RAML)	0.86

SEE: standard error estimate; R: correlation coefficient; ICC: intraclass correlation.

Validation of regression equations

The stature calculated for all the 200 subjects using the derived formulae was called estimated stature (Es). The stature measured using the PMCT was called observed stature (Os). The derived linear regression equations were validated using the ICC reliability statistics. A consistency of 71–89% was observed between Es and Os among the combined population while among males and females, the range of consistency was 70–89% and 79–88%, respectively. Similarly, the derived multiple regression equations were validated and the consistency range was 86–95% (Table 4).

Discussion

The stature of a living human being and a deceased is not the same due to proven postmortem changes like rigor mortis, etc. The stature is not the same when measured at different intervals during life due to the diurnal variations. There are physiological changes like spinal curvature changes due to the position the individual follows during life, degenerative changes in bones as a part of aging, etc., which plays a major role in determining the stature of an individual with life.⁹ This shows that there prevails an uncertainty in determining the stature of an individual both during life and after death as well. Several researchers concluded that there is a need for an application of adjustment factor while calculating the stature of cadavers to overcome such differences in height of living and dead.^10–12 Hence, forensic anthropologists/pathologists give an age range for the commingled skeletal remains considering the antemortem and postmortem changes.

The researchers mostly use instruments like vernier caliper, and inch tape less likely to collect the data required for analysis. In the present study, the 3D VRT along with an electronic cursor was used for data collection. Fernandes et al. concluded that the measurements taken by 3D VRT are reliable but not accurate.¹³ Therefore, the authors considered measuring all the measurements twice in VRT and utilized the mean values for analysis to avoid gross variations in lengths. The advantage of using VRT is that it has the benefit of real-time operation of object mass which enables to study the bones without altering the anatomy. There are few other studies that considered the VRT technique for data collection.^14–18 A study by Sakuma et al., compared the skull measurements using calipers and the 3D-VRT technique and observed no significant difference in obtained measurements.¹⁹ According to Giurazza et al. by using a CT scan, bone lengths can be measured accurately and a thickness of 1 mm is highly valuable for such a purpose.⁷ The slice thickness utilized in the present study is 1 mm and the authors of the present study based on this experience reiterate the comment of Giurazza et al.

After performing a thorough literature review, the authors were able to get only three papers published globally that have explored the stature estimation from the scapula using the 3D VRT technique, and the comparison of results (correlation values) is presented in Table 5. The differentiating feature between other studies and the present study is that all the measurements were studied in both the scapula bones, unlike other studies. The main findings of this study are LML had the maximum measurement and SMW had the minimum measurement among both males and females. A significant positive correlation observed between the stature and various scapular measurements on both sides among the combined and gender-wise population (p < 0.05) was observed. In the present study, the left-sided R values were higher than the right-sided values for males while among females same findings were observed except for LSMW and LTSL. The accuracy of prediction of stature using the derived formulae is less than 5 cm for all the variables studied irrespective of gender. The maximum and minimum confidence interval range noted was 2.5–4.14 and 3.62–5.5 for males, respectively, similarly 2.06–4.13 and 4.12–6.48 for females. The r values of the combined population are higher than the gender-wise population due to differences in the skeletal maturation physiology of the genders. The equations of the combined population will have significance in cases where the gender of the bone is difficult to find. In such a scenario, the stature of the bone could be concluded using the formula. The multiple regression equations had a higher R-value than the linear regression equation with a greater accuracy rate.

Table 5.

Comparison of the correlation coefficients between stature and scapular measurements from available literature among males and females.

	Torimitsu et al.²⁰(Japan)2011		Giurazza et al.⁷(Italy)2013		Zhang et al.(China)2016		Present study(Northern India)2021
	Male	Female	Male	Female	Male	Female	Male	Female
LSMW	—	—	—	—	0.47	0.35	0.66	0.61
LSML	—	—	—	—	0.44	0.35	0.75	0.58
LLSL	0.75	0.82	0.74	0.70	0.46	0.41	0.63	0.63
LLML	—	—	—	—	0.49	0.29	0.72	0.68
LTSL	0.59	0.64	0.51	0.48	0.46	0.40	0.70	0.67
LAML	—	—	—	—	0.46	0.49	0.62	0.67
RSMW	—	—	—	—	—	—	0.63	0.62
RSML	—	—	—	—	—	—	0.72	0.51
RLSL	0.76	0.81	—	—	—	—	0.58	0.61
RLML	—	—	—	—	—	—	0.68	0.67
RTSL	0.59	0.559	—	—	—	—	0.67	0.68
RAML	—	—	—	—	—	—	0.58	0.66

Torimitsu et al.²⁰ concluded that equations using LSL measurements from Japan population are reliable, while Zhang et al.²¹ who studied the Chinese population concluded LAM to be accurate for males and LML for females. The present Indian study concludes that for the males SML on both sides, and for females AML on the left side and TSL on the right side are reliable. The results of the present study in comparison with the other scapula studies had a lesser standard error (Table 6), the lowest being LSML (3.99) for males and RTSL for females (4.53). Campobasso CP et al.²² and Di Vella et al.²³ had extensively worked on the sex determination and stature estimation, respectively, from the scapula bone by taking the measurements directly from the bone rather than using any modality like PMCT. They measured and studied the lengths and breadths of the glenoid cavity and the distance between the acromion and coracoid processes individually. Their results were quite encouraging and proved that these measurements equally aid in the partial identification (stature and sex) of the individuals. The advantage of studying these measurements is that it does not require a fully intact scapula and would assist greatly when the skeletal remains are fragmented, including the scapula. The partial identification of an individual may be obtained from the available fragment of scapula bone consisting of the glenoid cavity or acromion and coracoid process. The study regarding partial identification utilizing the PMCT to measure the said measurements could be conducted in the future. Lastly, the author did not compare the observations with studies performed on any other regions considering establishing compiled data of four countries, namely, Japan, Italy, China, and India pertaining to the scapula.

Table 6.

Comparison of the standard error of estimate between stature and scapular measurements from available literature among males and females.

	Torimitsu et al.²⁰(Japan)2011		Zhang et al.²¹(China)2016		Present study(Northern India)2021
	Male	Female	Male	Female	Male	Female
LSMW	—	—	6.21	5.53	4.54	4.87
LSML	—	—	5.40	4.80	3.99	4.98
LLSL	3.79	4.37	7.21	5.15	4.70	4.78
LLML	—	—	6.01	4.63	4.17	4.49
LTSL	4.62	5.83	6.28	4.85	4.33	4.56
LAML	—	—	5.25	6.48	4.74	4.54
RSMW	—	—	—	—	4.70	4.83
RSML	—	—	—	—	4.23	5.27
RLSL	3.75	4.42	—	—	4.93	4.86
RLML	—	—	—	—	4.43	4.57
RTSL	4.62	6.31	—	—	4.51	4.53
RAML	—	—	—	—	4.94	4.60

Conclusion

To the best of our knowledge, this is the first study to derive regression equations for stature estimation using scapular measurements in the Indian population. In addition, this is the first-ever study globally to analyze all the possible measurements of the scapula on both sides; derive and validate the equations as well. The present study also has the benefit of deriving and validating the corresponding multiple regression equation for pooled population, males and females on both sides. This study concludes that scapular measurements may be used for calculating stature precisely in the contemporary North Indian population, in circumstances where more accurate skeletal elements, like intact long bones, are lacking for analysis. The derived equations may medicolegally serve the purpose of approximately estimating the stature range from the scapula bone since the standard error of the estimate is relatively less. However, more region, race, and ethnic groups-based studies analyzing the correlation between stature and scapular measurements are warranted in order to compile a list of helpful regression equations and compare the same for forensic investigations.

Footnotes

Ethics approval and consent to participate

The ethical clearance was given by the AIIMS institutional ethics committee vide IEC 577/02.11.2018, RP-29/2018. Consent forms were given and signed by the first-degree relatives (Legal Heir) of all subjects before participation.

Authors’ contribution

Karthi Vignesh Raj K: Conceptualization, Methodology, Validation, Formal analysis, Writing – original draft. G Gokul: Validation, Formal analysis, Writing – original draft. Abhishek Yadav: Conceptualization, Resources, Writing – Review & Editing, Supervision. Sudhir K Gupta: Resources, Supervision, and data. Swati Tyagi: Writing – Review & Editing, Supervision. Abilash Srinivasamurthy: Writing – Review & Editing. All authors read and approved the final manuscript.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study is a part of the project, Centre for Advanced Research and Excellence (CARE) in Virtual Autopsy, funded by the Indian Council of Medical Research (ICMR) Headquarters, New Delhi.

ORCID iD

Abhishek Yadav

References

Giurazza

Vescovo

Schena

, et al. Determination of stature from skeletal and skull measurements by CT scan evaluation. Forensic Sci Int 2012; 222: 398.e1–398.e9.

Zhang

Luo

Fan

, et al. Stature estimation from sternum length using computed tomography-volume rendering technique images of western Chinese. J Forensic Leg Med 2015; 35: 40–44.

Hishmat

Michiue

Sogawa

, et al. 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 Leg Med 2015; 129: 1173–1182.

Torimitsu

Makino

Saitoh

, et al. Stature estimation based on radial and ulnar lengths using three-dimensional images from multidetector computed tomography in a Japanese population. Leg Med 2014; 16: 181–186.

Karakas

Celbis

Harma

, et al. Total body height estimation using sacrum height in Anatolian Caucasians: multidetector computed tomography-based virtual anthropometry. Skeletal Radiol 2011; 40: 623–630.

Brough

Bennett

Morgan

, et al. Anthropological measurement of the juvenile clavicle using multi-detector computed tomography-affirming reliability. J Forensic Sci 2013; 58: 946–951.

Giurazza

Vescovo

Schena

, et al. Stature estimation from scapular measurements by CT scan evaluation in an Italian population. Leg Med 2013; 15: 202–208.

Giurazza

Schena

Del Vescovo

, et al. Sex determination from scapular length measurements by CT scans images in a Caucasian population. Conf Proc IEEE Eng Med Biol Soc 2013; 2013: 1632–1635.

Guharaj

Gupta

. Personal identity. In: Guhuraj

Gupta

(eds) Forensic medicine and toxicology. 3rd ed. Hyderabad, India: Universities Press, 2019, p. 77.

10.

Trotter

Gleser

. Estimation of stature from long bones of American Whites and Negroes. Am J Phys Anthropol 1952; 10: 463–514.

11.

De Mendonça

. Estimation of height from the length of long bones in a Portuguese adult population. Am J Phys Anthropol 2000; 112: 39–48.

12.

Mahakkanukrauh

Khanpetch

Prasitwattanseree

, et al. Stature estimation from long bone lengths in a Thai population. Forensic Sci Int 2011; 210: 279.e1–279.e7.

13.

Fernandes

Adamczyk

Poleti

, et al. Comparison between 3D volumetric rendering and multiplanar slices on the reliability of linear measurements on CBCT images: an in vitro study. J Appl Oral Sci 2015; 23: 56–63.

14.

Lai

Noor

MHM

Abdullah

. Stature estimation study based on pelvic and sacral morphometric among Malaysian population. Bull Nat Res Centre 2021; 45: 142.

15.

Koşar

Mİ

Gençer

Tetiker

, et al. Sex and stature estimation based on multidetector computed tomography imaging measurements of the sternum in Turkish population. Forensic Imaging 2022; 28: 200495.

16.

Kira

Chiba

Makino

, et al. Stature estimation by semi-automatic measurements of 3D CT images of the femur. Int J Legal Med 2022; 137: 359–377. https://doi.org/10.1007/s00414-022-02921-y

17.

Hishmat

Michiue

Sogawa

, et al. 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 2015; 129: 1173–1182.

18.

Lee

Gong

Hyun

, et al. Estimation of stature from femur length measured using computed tomography after the analysis of three-dimensional characteristics of femur bone in Korean cadavers. Int J Legal Med 2017; 131: 1355–1362.

19.

Sakuma

Ishii

Yamamoto

, et al. Application of postmortem 3D-CT facial reconstruction for personal identification. J Forensic Sci 2010; 55: 1624–1629.

20.

Torimitsu

Makino

Saitoh

, et al. Stature estimation in Japanese cadavers based on scapular measurements using multidetector computed tomography. Int J Legal Med 2015; 129: 211–218.

21.

Zhang

Cui

Luo

, et al. Estimation of stature and sex from scapular measurements by three-dimensional volume-rendering technique using in Chinese. Leg Med 2016; 21: 58–63.

22.

Campobasso

Di Vella

Introna

. Using scapular measurements in regression formulae for the estimation of stature. Boll Soc Ital Biol Sper 1998; 74: 75–82. PMID: 10874984.

23.

Di Vella

Campobasso

Dragone

, et al. Skeletal sex determination by scapular measurements. Boll Soc Ital Biol Sper 1994; 70: 299–305. PMID: 7748529.