The overall aim of this PhD-project was to advance the detection, analysis, and presentation to
laypersons of skull fracture in the field of forensic pathology. The advancement of computed tomography,
the increase in computational capability, and the rapid growth in 3D printing possibilities
provide tools that may propel forensic medicine from the autopsy- and microscopy investigations
of ancient times and into the 21st century. Studies I - IV focused on diagnosing blunt force
skull fracture with post-mortem computed tomography (PMCT), analysing skull fractures with
finite element analysis (FEA), and reporting the results to the court of law with the aid of 3D prints.
Study I aimed to provide a summary effect size for the sensitivity and specificity of PMCT for
skull fracture detection. Because individual studies on PMCT with varying sample sizes had reported
outcomes ranging from 0.50 to 1.00, a meta-analysis was needed. Study I demonstrated that
PMCT is a near perfect diagnostic test in terms of sensitivity and specificity i.e., finding fracture
in cases with an actual fracture and excluding fracture in cases with no actual fracture [1].
However, the existing studies included cases with extensive fractures caused by explosions, gunshots,
high-velocity blunt force traumas, etc. PMCT may be a better diagnostic tool in cases with
such trauma mechanisms than in cases with less forceful blunt force head trauma. If data from
PMCT are to be used to compare with results of FEA, or if PMCT is to be used as a screening tool
at medico-legal investigations, the diagnostic capabilities of PMCT for identifying small and subtle
fractures must also be determined.
Study II aimed to determine the diagnostic capabilities of PMCT for detection of individual fracture
lines in the skull and present this as sensitivity and specificity. A retrospective study compared
autopsy with PMCT and generated more than 5,000 data points. Study II found that even though
the presence and absence of a fracture system can be reliably diagnosed or excluded, the individual
fracture lines were difficult to diagnose, and inter-observer agreement was moderate [2].
Study III aimed to assess the feasibility of using subject-specific finite element (FE) head models
for FEA of blunt force skull fracture in adults in forensic pathology routine cases. Study III found
that FEA successfully predicted the occurrence of skull fracture in five of five cases, and correctly
predicted fracture pattern in three of five cases. FEA was sensitive to changes in kinematic assumptions.
FEA was found feasible in cases with linear fractures, a single head impact, and well
described circumstances. Subject-specific FEA may aid forensic pathologists in analysing skull
fracture and assessing the plausibility of proposed scenarios. The results should be used cautiously
though.
Study IV aimed to explore the effects of introducing 3D print as a demonstrative aid to forensic
pathologists providing expert testimony. Study IV found that the professional norms of forensic
pathologists are different to the norms of prosecutors, defence counsels, and judges. This difference
contributed to the court’s already existing problem of critically appraising e.g. statistical,
medical, and psychological statements. 3D prints enabled a quick overview, provided a mental
x
image of the fractures, and were perceived to be less emotionally confronting than autopsy photos.
However, physical 3D prints posed logistical challenges in distribution, demonstration, and archiving.
Virtual 3D models were perceived equally beneficial to 3D prints as the sense of touch
was of lesser importance than previous studies had speculated. Virtual 3D models were perceived
to cause less emotional impact than 3D prints and be easier to distribute, show, and store. Despite
the benefits of 3D demonstration, the court rarely needed the detailed information provided by 3D
prints, instead relying on the expert witness to assess proposed scenarios. 3D prints were therefore
rarely necessary.
In conclusion, this PhD-project demonstrated that while PMCT had a near perfect sensitivity and
specificity for skull fracture detection in a varied sample, the detection of individual fracture lines
was more difficult. PMCT must, for now, be supplemented by autopsy if all fracture lines are to
be visualised. PMCT data may be used for generation of subject-specific 3D models, which may
be used for FEA of skull fracture or 3D prints to demonstrate autopsy findings in court. Much
work remains before FEA and 3D printing can be implemented in routine forensic pathology but
the studies in this project highlights some of the drawbacks and difficulties, thus paving the way
for future research and implementation.