Analysis of Blunt Force Trauma in Human Cranial Bones: A distinction between perimortem and post-mortem fractures

(1)

Analysis of Blunt Force Trauma in Human

Cranial Bones:

A distinction between perimortem and

post-mortem fractures

Patricia Ribeiro, 11399872

36 ECTS

February 5

th

_{– July 30}

st

_{, 2018}

Submission date:

Presentation date:

Lay-out according to the guidelines of International Journal of Legal Medicine

Supervisor: Ignasi Galtés

1

Examiner: Rick van Rijn

2

Co-Examiner: Roelof-Jan Oostra

2

1 _{Institut de Medicina Legal i Ciències Forenses de Catalunya (IMLCFC), Barcelona. Ciutat de la Justícia, Gran}

Via de les Corts Catalanes Edifici G, 08075 Barcelona, Spain

(2)

Abstract

For a forensic anthropologist it is important to distinguish peri- from post-mortem fracture. These differences are crucial to determine the time when a fracture occured, reconstruct the events and provide possible scenarios of the fractures inflicted on the victim. There are studies that provide this information on long bones and ribs, but there is no literature on cranial bones. The main aim of this study is to obtain fracture pattern of cranial bones that might allow us to characterize a perimortem fracture. This was performed through macroscopic assessment of comminuted fractures between 123 fragments from autopsies, 100 fragments of experimental post-mortem fractured and 20 fragments of experimental fresh fractures.

The evaluation resulted in 6 traits associated with perimortem fractures: wave lines, peels, flake defects, fissures, crushed margins and bone scales. In addition, an algorithm was designed to predict the period of a fracture based on the frequency and significance of these traits.

Given the evolution of technology and the need to find a better method to evaluate cadavers in poor conditions, eight fragments from autopsy cases were submitted to 3D CT-scans. It was possible to distinctly observe 6 traits on a 3D CT scan: wave lines, bone scales, peel, bevelling, laminar breakage and bridge.

The preliminary results of the experimental fresh bones evaluation showed that there might be a correlation between the fracture morphology and the biomechanism of the injury. In this study, it is suggested that peels and flake defects occur due compressive forces. On the other hand, fissures might occur due to tensile stress.

(3)

Table of Figures

Fig 1: Pendulum mechanism: (1) arm of the pendulum, (2) hammer and (3) support for the craniums. 9 Fig 2:Communited fracture assembled to observe the general fracture patter on a fresh cranium (A); and a dry cranium (B). Black lines were added by the author to emphasize the pattern. ... 12

Fig 3:Hair on fresh bones, inserted in the suture ... 13

Fig 4: Wave lines on the outer table of a fresh bone. ... 13

Fig 5: Peel on the outer table of a fresh bone ... 14

Fig 6: Laminar breakage on the outer table of a fresh bone ... 14

Fig 7: Flake defect on the outer table of a fresh bones ... 15

Fig 8: Curved fissure on the outer table of a fresh bones ... 15

Fig 9: Crushed Margins on fresh cranial fragment ... 16

Fig 10: Bridge on fresh bones: (A) outer table of a cranium with two fractures; (B) inner table of the same cranium with two visible fractures. The different colors of the arrows show the same position of the fractures in both tables of the bone. ... 16

Fig 11: Bone scales on fresh bones: (A) coronal plane view; (B) sagittal plane view ... 17

Fig 12:Bevelling on fresh bones ... 17

Fig 13: Percentage of the traits' presence on three group samples ... 18

Fig 14: Schematic representation of the traits classification: PeriM – perimortem; PostM – post-mortem and the number between brackets present the number of fragments. ... 19

Fig 15: Volumetric reconstruction of the traits using the Toshiba Alquilin Primer (right) and the General Electrics Lightspeed Pro16 (left): wave lines (a), peel (b), laminar breakage (c), flake defect (d), bridge (e). bone scales (f) and bevelling (g) ... 22

(5)

1 Introduction

A forensic anthropologist usually works with skeletal remains and bodies in poor conditions such as burned, decomposed, mummified, saponified and bodies that are in a bad stage of preservation due to extrinsic factors. In these cases, one of the most challenging process is to determine the time period in which the injuries were produced[1]. The bones that constitute the human skeleton vary in thickness of the cortical and trabecular bones, but all of the them are composed by minerals, collagen and other proteins[2]. In living individuals, the presence of these components is extremely important to preserve the biomechanics of the bones. In case an individual is dead, and the bones are degraded, these components are no longer present. From a forensic point of view, these features allow the anthropologists to infer the timing of the injuries.

There are three distinct periods for the classification of a fracture: antemortem, perimortem and post-mortem. When a fracture is produced antemortem, the bone presents visible traits of healing. However, if there are no signs of healing processes it may indicate that it is either a peri- or post-mortem injury. The major difference is the presence or absence of the organic compounds. In the perimortem period, those components are present, allowing the bones to respond to an injury according to the strength and strain curve. This curve explains the elasticity and plasticity properties that a bone can withstand when a load is being applied to it [3]. Contrarily, in the post-mortem period the organic matrix of the bones is absent. For this reason, the bone break differently in these two periods, due to the presence or absence of the organic compounds that give the bone their plasticity and elasticity [2–9].

There are three methods to distinguish between perimortem and post-mortem. According to literature, the evaluation of anthropological cases can be performed in three different manners. Firstly, it can be done macroscopically, which is the most frequent and practical way of analysing the remains. Secondly, through histology, which is a method that is not frequently used due to the lack of information about it, but there are current studies focusing on the development of this method in forensic anthropology [10]. Finally, fracture evaluation can be done using imaging tools, such as X-Ray, Magnetic Resonance Imaging (MRI) and Computerize Tomography (CT) [11, 12].

(6)

According to Cappella et al (2014) the most common way to evaluate features in long bones are based on morphological and macroscopic characteristics such as: fracture pattern, fracture angle and the roughness of the fracture margins [1]. Moreover, anthropologists used to use the colour of the bone margins to differentiate peri- from post-mortem fractures. Nevertheless, it can be misleading since it is highly dependable on the climate and the type of soil, and it can change if the remains are moved [2, 6, 13–16]. Additionally, there are studies on blunt force trauma (BFT) fractures on long bones, where patterns have been established to distinguish peri- from post-mortem injuries. These studies are based on the macroscopic evaluation of the morphology of the fractures, correlating their findings with the mechanism of the injury [17, 18].

Although there is current literature that describes the timing of fracture on long bones, there is very little information about the cranium even though cranial injuries are very common in forensic cases. Cranial bones are composed of an outer and inner table of cortical bone, and in between there is tubercular bone. Within the cranium vault there is a variation of thicknesses and proportions of the trabecular bone. The mechanism of cranial injuries is generally explained by a combination of tension and compression loadings between the two layers. Its distinct geometry and architecture allows to infer a correlation between the biomechanics and the type of fracture [2].

Classically, skull fractures can be described as: linear, diastatic, depressed, stellate and comminuted. Linear and diastatic fractures occur when the energy of impact goes through the path that causes the least resistance. The difference is that diastatic fractures occur in the sutures. Depressed fractures are the collapse of the diploe. It highly associated with accumulated compressive forces on a small surface.

Stellate fractures are multiple radiating fractures that arise from the point of impact. They result from

a low velocity and heavy loading that it will cause the bone to bend. This type of fracture it’s more related with tensile forces. Lastly, comminuted fractures develop by the mechanism previously mentioned but results in extensive fragmentation of the bone, where reconstruction is often unlikely to occur. Usually, these fractures originate on the skull’s convexities where the central area is severely fragmented and concentric fractures are present away from the impact point [2].

(7)

In Forensic Anthropology, especially in bodies in poor conditions, it can be difficult to estimate the time when a cranial fracture occurred. This is due to the lack of soft tissues and the influence of taphonomic activity has when establishing a causal relationship. Moreover, the determination of fractures’ patterns on cranial bones may help the forensic anthropologist to determine when the fracture occurred. Unfortunately, there is a lack of information about this subject on cranial bones.

Nowadays, imaging tools are being studied and applied in the forensic field, such as 3D CT. In cases where the remains are in poor condition, it will allow the evaluation of the pattern and period when the fracture was performed. Furthermore, this technique will decrease handling of the specimens; it will allow the visualization of the skull in several planes which does not imply a destruction of the skull’s structure; and it is a method the can preserve the bone which sometimes is an evidence by itself [9, 19]. This way, future cases can be handled differently, giving a more accurate and scientific response and conclusion to anthropology cases in the forensic field. It will also allow forensic experts to understand the biomechanisms of fractures so that a reconstruction of events and a formulation of a plausible and possible scenario can be performed.

The aim of this study is to provide a distinct cranial perimortem fracture pattern, allowing to distinguish perimortem from post-mortem fracture on cranial bones. This will be performed by analysing fractures from autopsy cases and experimental cases using dry and fresh bones. We hypothesize that the experimental fresh bones will provide information about the biomechanism involved in the fracture pattern, when compared to the autopsy cases. Additionally, it is expected to observe a clear difference regarding the bones behaviour between the fresh and dry bones. All the evaluations were performed through macroscopic analysis of the fragments, so it was possible to provide a probabilistic tool for future cases. The fragments with more distinct traits will be submitted to a 3D CT-scan to infer the possibility of using this equipment as a tool in forensic anthropology. The distinction between peri-and post-mortem cranial injuries by blunt force trauma, as well as the understanding of the fracture itself, will make the forensic anthropologists a more reliable expert witness when in court. The admissibility of their testimony will fulfil the requirements requested by Daubert standard when the presented results are based on scientific method and accepted by the scientific community [20, 21].

(8)

2 Materials and Methods

This project includes both peri- and post-mortem fractured cranium samples. In total, 123 cranial bone fragments from 29 autopsy cases are being used. The bone’s sample came from forensic autopsies at the Institute for Legal Medicine and Forensic Science of Catalonia (IMLCFC), which were removed for complementary medico-legal investigation of the trauma. These fragments were collected from cases where high loading blunt trauma was applied to the cranium, such as falls from heights and vehicle accidents. The perimortem fracture patterns of autopsy cranial samples were compared with the pattern exhibited by experimentally fractured dry craniums.

The Medical Anatomy Department of the Autonomous University of Barcelona (UAB) provided 18 cases of dry unfractured craniums and 3 fresh unfractured human heads from people who donated their body to science. The bodies were preserved in an embalming mixture from PanReac AppliChem of ITW Reagents with phenol, ethanol, formaldehyde solution and glycerol. The study of bone behaviour will allow to understand the biomechanism of the fracture.

This study was approved by the Ethic Commission of Human and Animal Experimental Work (CEEAH) of the UAB, in compliance with the ethical regulations.

2.1 Fracture Reproduction

The craniums provided by UAB, both fresh and dry, were broken by custom-made pendulum. With this machine fractures were simulated so they could be compared to the fractures obtained from autopsies. The pendulum has a 1,50 m arm and a removable hammer of 8 Kg that was used to reproduce the fracture. This pendulum was based on the Charpy impact test [22]. The hammer has an impact surface of 50 cm2_{. (Fig 1). Furthermore, the mechanism has the option to brake, so it is possible to control the}

pendulum effect on the bones. To support the craniums, a surface was placed at the bottom of the apparatus parallel to it. This surface was not fixed to the apparatus, but it was supported in one side, so compression forces on the skull could be avoided.

Before each experimental reproduction of the fractures, all parameters were calculated and adjusted. On dry bones, the pendulum was released at a 90º angle. Given that the skin absorbs high levels of

(9)

energy provided during impact, the parameters were readjusted. For the experiments on fresh samples, the skin was first removed from the cranium and the pendulum was release a 135º angle.

For further calculations, the duration of impact was recorded with a high-speed camera. The duration of impact was considered to be from the moment the pendulum hits the cranium until there was no more dissipation of the energy from the impact, i.e. until the fracture was concluded.

Fig 1: Pendulum mechanism: (1) arm of the pendulum, (2) hammer and (3) support for the craniums.

2.2 Sample Preparation

Cranial fracture specimens, both from autopsy cases and experimental fresh bones, were further defleshed for subsequent examination. The defleshing method was based on a method described by Fenton [23]. The fragments were placed in a solution of water and degreasing detergent (5L:1cup) which was heated until 100 ºC. After reaching boiling point, the temperature was lowered to 80 ºC for 3-4 hours, depending on the amount of soft tissue and the bone itself, since the amount of grease varies between individuals. After the boiling process, the remaining soft tissues were carefully removed with chirurgical instruments. Finally, the bones were left to dry. In case the bones were still greasy, they were cooked again with 2.5% ammonia and water solution until the bones were grease-free, rinsed with water and set to dry.

2.3 Fracture Examination

Each fracture was examined as a whole, and each fragment was investigated individually. Due to these 1

2

(10)

two types of evaluation, not only the traits were visualized but it was also possible to observe the fractures’ progress. All bone fragments were assessed macroscopically, the margins were examined in more detail with the help of an amplifier (OPTIKA Microscopes). Next to the examination of perimortem specific traits, the traits were related to two more criteria: the position of trait regarding the bone – inner or outer table – and the position regarding the general pattern of the fracture – radiating or concentric. In case the latter criterion was unidentifiable, it was registered as unknown.

2.4 3D CT-scan Examination

Eights fragments, from autopsy cases, were carefully chosen to be examined by two 3D CT-scans. The chosen fragments had the most evident and pronounced.

Two CT machines were used to perform a 3D CT: The General Electric Lightspeed Pro 16 with 16 detectors, six images per second and each image is 0.625mm thick; and the Toshiba Aquilin Prime 80 with 80 detectors, with 60 images per second and each image is 0.5mm thick.

The scans were reconstructed through a commercial software, the 3D Slicer 4.8.1. This software allows to visualize the specimens in three planes: axial, sagittal and coronal, as well as the volumetric reconstruction.

2.5 Statistical Analysis

For this project, the frequency of each trait among each group sample was calculated. The data from each sample group was used to perform binary logistics regressions, along with Spearman’s correlation, to ascertain the possible associations between traits. Furthermore, a logistic model of multivariant analysis allowed to compare the fresh bones traits from the autopsy cases (perimortem cases) and dry bone traits (post-mortem cases). The statistical evaluation was performed through a Deducer JGR (Java Gui for R) Software version 1.7-9 and all p-values ≤ 0.05 were considered statistically significant. An algorithm for building a decision tree was used through WEKA 3.8 software. This tree was constructed through the classifier J48 using 10-fold cross validation.

(11)

3 Results

Table 1 gives an overview of the fractures obtained from autopsy cases categorized by age, sex and trauma circumstances. Most of the cases involved male victims (82.76%) and adults of the age 30-60 years old (51.72%). The trauma circumstances included vehicle collisions (65.52%) and falls from heights (34.48%). In total, 123 fragments were obtained from the autopsy cases.

Table 2 provides information regarding the sex of the experimental dry and fresh bone samples. The majority of donor sample were males (55.56% and 100% in dray and fresh bones, respectively) and adults between the age of 31-60 years old in dry bones and 67-70 years old in fresh bones. The dry group provided 100 fragments that were compared to the fragments from the autopsy cases. The experimental fresh cases provided 20 fragments that were later studied to understand the biomechanical implication of the fracture traits.

Table 1: Sample distribution of autopsy cases regarding age, sex and circumstances of death

Cases Percentage (%)

Age Young Adults (18-30 years) 5 17.24

Adults (31-60 years) 15 51.72

Elderly (≥61 years) 3 10.34

Unknown 6 20.70

Sex Female 5 17.24

Male 24 82.76

Trauma Circunstances Vehicle colission 19 65.52

Fall from heights 10 34.48

Table 2: Sample distribution of experimental dry bones and fresh bones regarding sex and age.

Experimental Dry Bones Experimental Fresh Bones Cases Percentage (%) Cases Percentage (%)

Age Adults (31-60 years) 14 77.78 - -

Elderly (≥ 61 years) 4 22.22 3 100

Sex Female 8 44.44 0 0

(12)

3.1 General Fracture Pattern

The force of impact on the experimental dry bones was 1250-4376N. Firstly, the velocity was calculated through the conservation of mechanical energy. This parameter was assessed throught the angle in which the pendulum was released and the length of its respecive arm. Secondly, through the time and velocity of the impact, as well as the mass of the pensdulum, the force was estimated. The impact force from the autopsy cases was impossible to infer since the specimens came from real life accidents. In order to assess the general fracture pattern, two craniums were reconstructed manually: one from an autopsy case and one from the dry bone sample group.

On the fresh cranium, there is a clear defined pattern formed by radiating fractures which originate from the impact point and concentric fractures (Fig 2a). In a few cases, small bits of hair were found between layers of cortical bone, with prominent presence in between the cranial sutures. (Fig 3)

On the other hand,, in dry bones there is not a specific pattern oberverd. The Fig 2b illustrates that dry bones tend to break randomly.

Fig 2:Communited fracture assembled to observe the general fracture patter on a fresh cranium (A); and a dry cranium (B).

Black lines were added by the author to emphasize the pattern.

(13)

Fig 3:Hair on fresh bones, inserted in the suture

3.2 Definition of the traits

The assessment of the 123 fragments from the autopsy cases presented a set of distinct traits: wave lines, peels, laminar breakage, flake defects, fissures, crushed margins, bridge, bone scales and bevelling. All the traits appeared either on the outer table or on the inner table of the craniums.

Wave Lines

Wave lines are smooth and small undulation that occur in the smooth edges of the fracture. They

resemble a wave with a long slope going up and a steep slope down. (Fig 4).

Fig 4: Wave lines on the outer table of a fresh bone.

Peel

(14)

Fig 5: Peel on the outer table of a fresh bone

Laminar Breakage

Laminar breakage is a missing superficial thin layer of cortical bone, which is not limited to margins

but covers a larger area on the fragment. It is the imprint of a peel on the adjacent fragment. (Fig 6)

Fig 6: Laminar breakage on the outer table of a fresh bone

Flake Defects

A flake is a superficial piece of cortical bone missing. It is a feature that can be easily lost in surrounding tissue. The imprint on the adjacent fragment is called a flake defect, which in this study can only be found on the margins. (Fig 7)

(15)

Fig 7: Flake defect on the outer table of a fresh bones

Fissures

A fissure is a thin crack that affects only the external or the internal cortical bone. According to the

shape and the position to the margins of the fragment it can be classified as: curved, irregular, parallel or perpendicular. (Fig 8)

Fig 8: Curved fissure on the outer table of a fresh bones

Crushed Margins

Crushed marginsare damaged cortical edges as a result of a fracture. It occurs on the margin of the fracture and can be identified by a thin discoloured line along the margins. It can be crushed upwards or downwards, depending the applied forces. It is a vulnerable trait that can easily be destroyed if not handled carefully. (Fig 9)

(16)

Fig 9: Crushed Margins on fresh cranial fragment

Bridge

When both cortical layers are fractured but the fragments are kept together by the trabecular bone layer it is called a bridge. (Fig 10)

Fig 10: Bridge on fresh bones: (A) outer table of a cranium with two fractures; (B) inner table of the same cranium with two

visible fractures. The different colors of the arrows show the same position of the fractures in both tables of the bone.

Bone Scales

Bone scales are small pieces of cortical bone still attached to the main fragment and often present plastic

deformation. They are very fragile features, resembling fish scales. The bone scales can appear in groups or individually and can only on be observed on the cortical bone (Fig 11).

B

A

(17)

Fig 11: Bone scales on fresh bones: (A) coronal plane view; (B) sagittal plane view

Bevelling

Bevelling is a gradual loss of trabecular and cortical bone on the edge of the fracture, leaving a sloped

edge cut (Fig 12).

Fig 12:Bevelling on fresh bones

3.4 Distribution of the traits

Distribution of the traits was compared between autopsy cases, experimental dry and fresh bones. (Fig 13) presents the distribution of the nine traits on each group sample: autopsy cases, experimental dry and fresh bones.

Among the samples from autopsy cases the most common trait was bevelling (78%) followed by laminar breakage (59.3%). The least common traits were bridges (1.6%) and bone scales (17.9%). In experimental dry bones, the most common traits were bevelling (81%) and laminar breakage (70%). Five traits were not found in this sample group: wave lines, flake defects, crushed margins, bridges and bone scales.

In experimental fresh bones, the majority of these traits were present. The two most common traits were

(18)

bevelling (70%) and peels (70%), followed by fissures (50%). The only difference between this sample group and the autopsy samples, was the absence of crushed margins.

Fig 13: Percentage of the traits' presence on three group samples

Table 3 shows the relationship between the traits and the peri- and post-mortem period. Through the Spearman’s correlation and the p-value, it was possible to relate the traits to the perimortem period (fresh condition). Those traits were: wave lines, peels, flake defects, fissures, crushed margins and bone scales (p-value <005). Even though peels are statistically significant and is correlated with the perimortem fracture, it has the lowest likelihood ratio (LR) present in all the correlated traits. (LR=4.18), whereas the highest LR was calculated for wave lines (LR=57.85). The likelihood ratio assessment allowed to assert that is approximately 58 times more probable to find wave lines in a perimortem fracture, rather than in a post-mortem fracture.

From the number of fragments of fresh and dry bones (n=223), 192 were correctly classified as perimortem and 92 were correctly classified as post-mortem, this gives this study an accuracy of 87%. This accuracy was calculated using the WEKA software.

Table 3:Logistic regression using traits and peri-and post-mortem period with p-value, Spearman correlation and likelihood ratio

Traits p-value Spearman Correlation (rs) Likelihood

Wave Lines < 0.001 0.434 57.85 Peel 0.043 0.136 4.18 Laminar Breakage 0.100 -0.11 2.74 Flake Defects <0.001 0.389 46.91 Fissures <0.001 0.297 21.38 Crushed Margins < 0.001 0.342 36.57 34.1 _30.9 59.3 28.5 32.5 22.8 1.6 17.9 78 0 19 70 0 8 0 0 0 81 25 70 30 35 50 0 15 25 70 0 10 20 30 40 50 60 70 80 90

Wave Lines Peels Laminar Breakage

Flake Defects

Fissures Crushed Margins

Bridge Bone Scales Bevelling

Perc en ta ge (% ) Traits

(19)

Bridge 0.202 0.086 2.4

Bone Scales <0.001 0.298 28.13

Bevelling 0.590 -0.036 0.295

The following decision tree is a simple way to analyse the presence of each trait in the perimortem period. Furthermore, it allows to predict the presence of the mentioned traits in future cases of unknown samples. For this prediction the only perimortem trait that was not used was peels, this is justifiable since its p-value (0.049) is close to the significant level (0.05).

The numbers between brackets is the number of fragments from the sample that had the given profile. In 113 fragments from the sample, all five traits were absent (113/21) which indicates that the fracture was classified as post-mortem. Given that this algorithm has 87% of accuracy, it is estimated that 21 cases were wrongly classified. Same reasoning is performed in case only fissures are present (24/8). (Fig 14Fig 14: Schematic representation of the traits classification: PeriM – perimortem; PostM – post-mortem and the number between brackets present the number of fragments.

)

Fig 14: Schematic representation of the traits classification: PeriM – perimortem; PostM – post-mortem and the number between brackets present the number of fragments.

Table 4 represents the correlation between the traits and the fracture pattern. This pattern was defined in inner or outer table of the cranium and radiating or concentric fractures.Wave lines, peels and flake defects were positive correlated with the outer table of the cranium. Laminar breakage and bone scales

(20)

were positive correlated with both inner and outer table, but both traits were more significant on the outer table of the skull. Fissures were the only trait correlated positively with the inner table of the cranium.

The same table provides information about the position of the trait regarding to the fracture. Peels, fissures and flake defects were correlated to the radiating fracture. Wave lines, laminar breakage, bridge and bone scales were only present on the radiating fracture. Since there is no information on the presence of these traits on the concentric fracture, the software wasn’t able to give a statistical significance. Moreover, all bridges were present on both layers of the cranium bone and the radiating fractures. Bevelling was not statistically significant on either positions, and crushed margins were not present on the fragments of this sample group.

Table 4: Correlation between the traits and the position regarding the bone layer and the fracture general pattern on experimental fresh bones.

Traits Position regarding the bone layer Position regarding the fracture

Inner table Outer Table Both Concentric Radiating Both

Wave Lines 0.397 0.866 - - All* -

p-value 0.083 <0.001 - - -

Peels 0.275 0.480 0.327 0.327 0.592 0.150

p-value 0.241 0.032 0.159 0.159 0.006 0.527

Laminar Breakage 0.509 0.642 0.350 - All* -

p-value 0.022 0.002 0.130 - - -

Fissures 0.489 0.420 0.500 0.333 0.655 0.333

p-value 0.009 0.065 0.025 0.151 0.002 0.151

Flake Defect - All* - 0.313 0.787 0.313

p-value - - - 0.180 <0.001 0.180

Crushed Margins - - - -

p-value - - - -

Bridge - - All* - All* -

p-value - - - -

Bone Scales 0.577 0.728 - - All* -

p-value 0.008 <0.001 - - - -

Bevelling 0.275 0.378 - 0.275 0.429 0.378

p-value 0.100 0.059 - 0.241 0.059 0.100 *all traits were present in this category.

3.5 3D CT-scan

Computerized tomography (CT) was performed with two different equipment, Toshiba Alquilin Primer and General Electrics Lightspeed Pro16. Images were processed to obtain a volumetric reconstruction on both CT-scans.

(21)

especially useful to differentiate between cortical and trabecular bone: wave lines, peel, laminar breakage, bridge and bevelling are visible. In both CT-scans bone scales were visible. The difference is the low image quality can cause some bone scales can go unnoticed in the General Electrics CT-scan, especially if the trait is not clearly distinguishable. Flake defect is a trait that can be difficult to observe unless there is access to the software. Some knowledge in image manipulation in the software, such as contrast and the ability to rotate the volumetric reconstruction, may help to better observe this feature. Crushed margins and fissures are not possible to observe (Fig 15).

(22)

Fig 15: Volumetric reconstruction of the traits using the Toshiba Alquilin Primer (right) and the General Electrics Lightspeed Pro16 (left): wave lines (a), peel (b), laminar breakage (c), flake defect (d), bridge (e). bone scales (f) and bevelling (g) a b c d e f g

(23)

4 Discussion

In the forensic field, the condition of the body determines how the autopsy is going to be conducted. The forensic pathologist mainly evaluates the evidences in the soft tissues to reconstruct the trauma circumstances, but the fracture pattern also gives clues about the mechanism implicated in the injury. Cranial trauma is one of the most common injuries in cases of death related traumas [24]. In cases the cranium appears in poor conditions, incomplete and/or damaged, the forensic anthropologist may help the pathologist to analyse the fracture pattern and to infer all these data. A cranium in those conditions make the reconstruction and the analysis of the fracture even more complex [2]. This complexity also includes the determination of the period of the injury.

Timing injury includes the classification of a peri- or a post-mortem trauma, which is one of the most problematic issue in the field of forensic anthropology. The importance of data fractures is related to the circumstances of death, as well as reconstruction of possible scenarios surrounding the time of death [4]. In literature, there are different criteria to distinguish peri- from post-mortem fractures. These criteria usually are based on macroscopic assessment of the fractures, as is the fastest way to realize the diagnostic. The present information on timing fractures is based mainly on colour, angle and edges of the fracture [2]. Recently, Scheirs et al (2016) compared peri- with post-mortem trauma in long bones in order to identify a distinct pattern. The authors concluded that there are specific traits that can help determine the time period of the inflicted fracture which is represented by original traits: layered breakage, flakes and flake defects, bone scales, wave lines and crushed margins [17]. On contrast, regarding cranial perimortem fractures, there is no literature concerning fracture patterns on cranial bones correlated to the period that the injury occurred.

It is well known that bones in a fresh and dry state break differently. When a fresh bone is submitted to a load it will first bend due to its elastic properties. This property allows the bones to return to the original form when the load is removed. Secondly, in case the load continues to be applied, the behaviour of the bone changes according to its condition. The fresh bone will keep bending and if the load is removed, the shape of the bone will be altered and not change to its original form. In case of a dry bone due to degradation of collagen fibres, it will break little or shortly after the yield point. This

(24)

condition is directly related to the almost or absolute absence of plastic deformation [2, 25, 26]. Resulting fracture morphology will depend on the structure and shape of the bone.

Cranial fractures have their own and unique characteristics such as its spherical shape, the presence of sutures and the structure of cortical and trabecular bone layers conditionate the fracture pattern [2]. Hart

et al (2005) explained the behaviour of a cranial fracture in cases of blunt force trauma. According to

these authors, in cases where high force is applied to the skull, there are three type of fractures. The primary fracture is the impact of the blunt force object. Secondly, the dissipation of the energy can generate secondary (radiating) and tertiary (concentric) fractures. This type of behaviour gives a spider-web appearance to the injury [2, 27]. Given the carachteristics of the cranium, it is hypothesised that the perimortem pattern of the cranial fractures will be different from the pattern obtained in long bones.

A mechanical pendulum was used to simulate cranial fractures in dry and fresh bones. By means of our experimental analysis reproducing in fresh cranial bones we were able to reproduce the spider web pattern. The results showed the impact site on the centre of the parietal bone, radiating fractures diverged from the impact site and concentric fractures appeared surrounding this same area. In case the injury includes the sutures, this pattern will also include these types of fractures, but the appearance will be slightly different. On the experimental fresh bones, it was also possible to reproduce radiating and concentric fractures. On the three specimens, the fracture went across the suture. These results were expected since the bodies from which the heads were collected, used to belong to X years old individuals. At this age the coronal and sagittal sutures are starting or are completely fused [28]. It is well known that the sutures can have an impact on the morphology of the fracture [2]. The general pattern of a fracture will vary, depending if the sutures are open, closed or completely fused. There is a higher probability that the energy will disperse throughout the sutures when the they are less unified. Otherwise, the fracture will propagate across these fixed joints [2]. Furthermore, Fleming et al (2013) stated that sutures are the weakest points on the cranium, which makes it easier to break. Moreover, in post-mortem cases the degradation of the collagen fibres and connective tissue occurs, making these structures even more predisposed to separate [9].

(25)

Moreover, on dry bones the results showed that this general pattern was extremely irregular. It was possible to observed how the energy of the impact dissipated through the skull and its random breakage path. In these cases, the sutures still present the path of least resistance and the fracture is most likely to evolve through it. This also contributes to the randomness of the injury [2, 9]. Even though the impact site was purposefully established on the parietal bone, other parts of the skull were involved during breakage. In all these specimens, the impact energy caused squamosal suture to completely detach itself from the skull. Furthermore, on the young adults’ (31-40 years old) specimens had the sagittal and coronal sutures closed, but not fused. In these cases, the impact induced the fracture to occur throughout the sutures causing them to open. In older specimens (50- <61 years old), the sagittal and coronal sutures were fused or partially fused, and the impact energy caused the fracture to go across the suture [28].

Although this general pattern can help to establish the differential diagnosis between peri-and post-mortem fractures, when it comes to bone fragments or remains in intermediate states of desiccation, this distinction will be more complex. In the literature there are no studies describing specific traits for differentiate between peri- and post-mortem cranial fractures. The ability to make this distinction would help the forensic anthropologist to improve the assessment of a blunt trauma to the cranium, especially in the mentioned cases. The results of this study allow us to identify nine traits in fresh fractures: wave lines, peels, laminar breakage, flake defects, fissures, crushed margins, bridge, bone scales and bevelling. In comparison with the dry craniums, it was possible to develop an algorithm to preform distinguish peri- from post-mortem fractures. This algorithm is a probabilistic approach presented in the form of a decision tree, in which was based on perimortem traits. The order of appearance of those traits is based on their frequencies on the perimortem group sample: wave lines, flake defects, crushed margins, bone scales and fissures.

When involved in judicial activities, not only is important to evaluate the evidences in the laboratory, but also to prepare to present the results to a judge. It is extremely important to provide reliable scientific testimony, so it can be admissible in court. The admissibility of an expert's testimony is to meet Daubert's criteria. One of these criteria is the necessity of a scientific method behind the data analysis [20, 21]. The algorithm designed in this study to predict the period of a fracture is the first tool that can

(26)

be used in future case of forensic anthropology. It will allow to assess cranial fractures by blunt force trauma, so it is possible to provide the timing of this type of injuries.

The results of the macroscopic evaluation of the perimortem traits were similar from the ones observed in long bones fractures. When comparing these traits with the ones reported by Scheirs et al (2016) it is possible to observe that wave lines, flake defects, crushed margins and bone scales were present in both types of bones. In contrast, fissures were unique to the cranial bones whereas layered breakage was only found on the long bones. The latter trait is distinctly different from laminar breakage since, layered breakage are layers in the cortical bone in the compression site in long bones [17] while laminar breakage is a thin missing superficial thin layer of cortical bone that it is the imprint of a peel. Scheirs

et al (2016) concluded that wave lines, flake defects and bone scales were correlated to the bone

architecture and to the presence of musculoskeletal activity on the long bone. The latter doesn’t affect the cranium, in this study we tried to reproduce the fractures on experimental fresh craniums to make that correlation for better understanding of the bone behaviour when submitted to a blunt force trauma. Until now there is no literature about the relationship between the biomechanics and the fracture morphology on cranial bones.

Imaging tests are increasingly present in forensic anthropology [9, 19, 29]. This is a method particularly used when human remains are in poor condition, such as saponified or mummified bodies. This study also approaches the analysis of cranial traits fractures using two different 3D CT-scanning machines: Toshiba Alquilin Prime with 80 detectors and General Electrics Lightspeed Pro 16 with 16 detectors. We tested how these two distinct equipment presented the traits and the respective differences regarding the image quality. The obtained result demonstrated that wave lines, peel, laminar breakage, flake defects, bridge, bone scale and bevelling were possible to observe in a CT-scan. The Toshiba, being one of the most recent equipment on the market has the technology to perform a better capture of the fragments details.

In the 3D CT scan, there is access to bidimensional CT scan in three different plans: axial, sagittal and coronal plane, which allows a better understanding of the traits. It also facilitates the observation of

(27)

certain characteristics of the fragment that is impossible to do in an axial CT-scan such as: peel, flake defect and laminar breakage. Even though the flake defects were difficult to observe on both 3D CT-scans, the Toshiba Prime Aquilin was able to show the trait with more quality than the General Electrics. The difficulty of observing the crushed margins and fissures in the volumetric reconstructed fragments is explained by the size of these traits and the sensitivity of the CT-scan.

Moreover, our results allowed us to make a preliminary approach about the biomechanics of the cranial fracture in blunt force trauma and specifically of the reported perimortem traits. -by means if experimental approach we have correlated the traits with the spider web fractures pattern and, specifically with the presence of radiating and concentric fractures. It is important to highlight that fresh specimens were used after the removal of the skin. There are studies that state that skin can absorb high amounts of energy and to tear this barrier is necessary to apply high loading [30, 31]. Moreover, they were fixed with an embalming solution of ethanol, glycerol, phenol and formaldehyde. There is a controversial discussion about the influence of the embalming solution on the bone behaviour. Despite the fact that these authors stated that formaldehyde and ethanol fixation methods can make the bone more brittle, they also stated that the differences were not significant enough to prove this concept [32– 36]. Given the presence of the skin and internal organs, we attempted to increase the force of impact by increasing the angle of which the pendulum was released. Unfortunately, it was not possible to calculate the exact force of impact due to the impossibility of recording the impact time. For future studies, it is necessary to adjust record the impact in several perspectives, so it is possible to assess the beginning and end of the fracture to establish the time of impact.

According to Sauer et al (1997), when the cranium is hit with a blunt object the bone bends inwards: the outer table undergoes compressive forces and the inner table undergoes tensile forces [37]. Hart et

al (2005) also explains that, in the surrounding areas of the impact, there is an increase of compressive

forces on the outer table and tensile forces on the inner table, forming the radiating fractures [27]. After these fractures are produced there is an inversion in the applied forces in the skull. The concentric fractures progress goes from the outer table, where tensile forces applied, to the inner table, where compressive forces are applied [37].

(28)

The results show that peels were statistically significant on the outer table of the radiating fracture. All the flake defects were present on the outer table of the skull. Given that there was no comparison on the trait regarding its position on the bone layer, the software could not provide information about the statistical significance. On contrast, this trait was also significant on the radiating fracture. Taking into account the location of the trait, it is assumable that it occurs due to compressive forces. These results are according to the conclusions of Scheirs et al (2016) where flake defects were attributed to the compression side of the fracture on long bones. Moreover, this experimental component of the study showed that fissures were more significant on the inner table of the radiating fracture. According to Hart et al (2005), this trait occurred when the bones undergoes tensile stress.

Wave lines, laminar breakage and bone scales were more statistically significant on the outer table of the skull. Moreover, all these traits were present on the radiating fracture. Although there is no statistic prove that these traits are always present on this position of the fracture, the preliminary results show that in this sample, these traits occur due to compressive forces on the cranium. Furthermore, bevelling is a trait that does not show any statistical significance in either position on the cranium. Nonetheless, it is almost significant on the outer table of the radiating fracture. This suggests that the way that the bone tends to break from the inner table to the outer table of the injury, creating a sloped edge. With the information available, we hypothesise that there is a possibility that bevelling occurs due to compressive forces. It is important to notice that this lack of statistical information might be related to the sample size. Further studies need to be performed about the biomechanism of the cranial fracture by blunt force trauma to be able to provide more information on these traits.

(29)

5 Conclusion

Perimortem cranial fractures could be defined with nine distinct macroscopic traits were found during the macroscopical evaluation: wave lines, peels, laminar breakage, flake defect, fissures, crushed margins, bridge, bone scales and bevelling. From there traits we present an algorithm based on the presences and absence of wave lines, flake defects, crushed margins, bone scales and fissures, which allow us to distinguish between peri- and post-mortem cranial fractures. The fragments that were submitted to 3D-CT showed that most traits were visible on the CT-scans except for crushed margins and fissures.

The experimental fresh bones were used as a preliminary study to be able to understand the biomechanism of a perimortem fracture. Peels and flake defects were attributed to compressive forces, while fissures were related to tensile stress. One limitation of this study was the number of samples used on experimental part of the project. By increasing the number of samples, more information will be performed to relate the fracture morphology and its biomechanics. This would allow the forensic anthropologists to understand the behaviour of cranial bones when submitted to blunt force trauma and, consequently, comprehend the manner of death more in depth.

6 Acknowledgements

I would like to show my deepest appreciation to my supervisor, Ignasi Galtés, not only for the guidance and knowledge, but also for the enthusiastic confidence that he shared with me throughout this project. I would like to thank Xavier Jordana for the essential help with the statistics test and Marisa Ortega and Gabriel Font for helping me with the experimental part of the study. A special thanks to Sarah Scheirs and Rick van Rijn for the advices and constructive suggestions during this research, and to Roelof-Jan Oostra for taking the role as my co-examiner in such short notice.

. I also want to show my appreciation to João Freitas for helping me regarding the physics behind the pendulum’s movement.

(30)

I wish to extend my gratitude to all the technicians from IMCL for helping me with the sample collection and to UAB to donate the samples for the experimental part of the project.

Finally, I would like to thank my family and friends for the unquestionable support during this research project.

(31)

8 References

1. Cappella A, Sc B, Amadasi A, et al (2014) The Difficult Task of Assessing Perimortem and Postmortem Fractures on the Skeleton : A Blind Text on 210 Fractures of Known Origin *. 59:1598–1601 . doi: 10.1111/1556-4029.12539

2. Vicki L (2014) Broken Bones: Anthropological Analysis of Blunt Force Trauma, 2nd ed. Charles C Thomas, Illinois

3. Wescott DJ (2013) Biomechanics of Bone Trauma. Encycl Forensic Sci 83–88 . doi: 10.1016/B978-0-12-382165-2.00015-5

4. Houck MM (2012) A Companion to Forensic Anthropology. Wiley-Blackwell

5. Işcan MY, Loth SR (1986) Determination of age from the sternal rib in white females: a test of the phase method. J Forensic Sci 31:990–9

6. Pokines JT, Symes SA (1996) Forensic Taphonomy

7. Berryman HE, Symes SA Recognizing Gunshot and Blunt Cranial Trauma Through Fracture Interpretation. In: Forensic Osteology. pp 333–352

8. Lindström A-C, Hung N, Duncan W, Kieser J (2014) Detection of gunshot residues (GSR) on a self-inflicted gunshot wound. Pathology 46:260–263 . doi: 10.1097/PAT.0000000000000083 9. Fleming-Farrell D, Michailidis K, Karantanas A, et al (2013) Virtual assessment of

perimortem and postmortem blunt force cranial trauma. Forensic Sci Int 229:162.e1-162.e6 . doi: 10.1016/j.forsciint.2013.03.032

10. Kranioti E (2015) Forensic investigation of cranial injuries due to blunt force trauma: current best practice. Res Reports Forensic Med Sci 5:25 . doi: 10.2147/RRFMS.S70423

11. Thali M (2011) Virtuelle Autopsie (Virtopsy) in der Forensik: Vom Skalpell zum Scanner. Pathologe 32:292–295 . doi: 10.1007/s00292-011-1520-5

12. Pechníková M, Porta D, Cattaneo C (2011) Distinguishing between perimortem and postmortem fractures: Are osteons of any help? Int J Legal Med 125:591–595 . doi: 10.1007/s00414-011-0570-9

(32)

14. Quatrehommea G, Yagar M (1997) hensic Science Interaatiml Postmortem skeletal lesions. 89:155–165

15. Wieberg DAM, Wescott DJ (2008) Estimating the timing of long bone fractures: Correlation between the postmortem interval, bone moisture content, and blunt force trauma fracture characteristics. J Forensic Sci 53:1028–1034 . doi: 10.1111/j.1556-4029.2008.00801.x 16. Coelho L, Cardoso HF V (2013) Timing of blunt force injuries in long bones: The effects of

the environment, PMI length and human surrogate model. Forensic Sci Int 233:230–237 . doi: 10.1016/j.forsciint.2013.09.022

17. Scheirs S, Malgosa A, Sanchez-Molina D, et al (2017) New insights in the analysis of blunt force trauma in human bones. Preliminary results. Int J Legal Med 131:867–875 . doi: 10.1007/s00414-016-1514-1

18. Scheirs S, Langenhorst W, Malgosa A, et al (2018) Perimortem fracture pattern in ribs by blunt force trauma

19. Burke MP (2012) Forensic Pathology of Fractures and Mechanisms of Injury - Postmortem CT Scanning. CRC Press

20. Grivas CR, Komar DA (2008) Kumho, Daubert, and the Nature of Scientific Inquiry: Implications for Forensic Anthropology. J Forensic Sci 53:771–776 . doi: 10.1111/j.1556-4029.2008.00771.x

21. Lesciotto KM (2015) The Impact of Daubert on the Admissibility of Forensic Anthropology Expert Testimony. J Forensic Sci 60:549–555 . doi: 10.1111/1556-4029.12740

22. Lucon E (2016) Experimental Assessment of the Equivalent Strain Rate for an Instrumented Charpy Test. 121: . doi: 10.6028/jres.121.007

23. Fenton TW, Birkby WH, Cornelison J (2003) A fast and safe non-bleaching method for forensic skeletal preparation. J Forensic Sci 48:274–276 . doi: 10.1520/JFS2002034

24. Li M, Zhao Z, Yu G, Zhang J (2016) Epidemiology of Traumatic Brain Injury over the world: A Systematic Review. Austin Neurol Neurosci 1:

(33)

26. Hannon P, Knapp K (2006) Forensic Biomechanics

27. Hart GO (2005) Fracture pattern interpretation in the skull: Differentiating blunt force from ballistics trauma using concentric fractures. J Forensic Sci 50:1276–1281 . doi:

10.1520/JFS2004219

28. Schünke M, M. Ross L, Schulte E, et al (2007) Thieme Atlas of Anatomy: Head and Neuroanatomy, First. Stuttgart

29. Sieswerda-Hoogendoorn T, Soerdjbalie-Maikoe V, De Bakker H, Van Rijn RR Postmortem CT compared to autopsy in children; concordance in a forensic setting

30. Whittle K, Kieser ÆJ, Ichim ÆI (2008) The biomechanical modelling of non-ballistic skin wounding : blunt-force injury. 33–39 . doi: 10.1007/s12024-007-0029-y

31. Sharkey EJ, Cassidy M, Brady J, et al (2012) Investigation of the force associated with the formation of lacerations and skull fractures. Int J Legal Med 126:835–844 . doi:

10.1007/s00414-011-0608-z

32. Thavarajah R, Mudimbaimannar VK, Elizabeth J, et al (2012) Chemical and physical basics of routine formaldehyde fixation. J Oral Maxillofac Pathol 16:400–5 . doi:

10.4103/0973-029X.102496

33. Hammer N, Voigt C, Werner M, et al (2014) Ethanol and Formaldehyde Fixation Irreversibly Alter Bones’ Organic Matrix. J Mech Behav Biomed Mater 29:252–258

34. Brenner E (2014) Human body preservation - old and new techniques. J Anat 224:316–344 . doi: 10.1111/joa.12160

35. Stefan U, Michael B, Werner S (2010) Effects of three different preservation methods on the mechanical properties of human and bovine cortical bone. Bone 47:1048–1053 . doi:

10.1016/j.bone.2010.08.012

36. Öhman C, Dall’Ara E, Baleani M, et al (2008) The effects of embalming using a 4% formalin solution on the compressive mechanical properties of human cortical bone. Clin Biomech 23:1294–1298 . doi: 10.1016/j.clinbiomech.2008.07.007

37. Sauer NJ (1997) The Timing of Injuries and Manner of Death: Distinguishing Among Antemortem, Perimortem and Post-mortem Trauma. In: Reichs KJ (ed) Forensic Osteology:

(34)

Advances in the Identification of Human Remains, 2nd ed. Charles C Thomas, Springfield, Illinois, pp 321–352

Analysis of Blunt Force Trauma in Human Cranial Bones: A distinction between perimortem and post-mortem fractures