An in-depth evaluation of outlier trajectories present in impact patterns

Abstract

Bloodstain pattern analysis (BPA) is an important field of forensic science that helps with crime scene reconstruction. One particular aspect of BPA is area of origin determination: this is the calculation of the 3D location of a blood source that was impacted to create a pattern of recognizable characteristics on a target surface. Previous research has helped optimize methods to calculate the area of origin using the length and width of multiple stains in an impact pattern. There are certain stains that are appropriate for origin determination, however even when the proper stains are selected, some still do not point back to the true blood source. These were termed outliers, and were the focus of this research. This research investigated a suitable distance from the true origin in order to consider a trajectory an outlier; if the trajectory was farther than the decided distance, it was considered an outlier. It also looked into if impact patterns generated outliers consistently, as well as if those outliers were recognizable based on the stain’s morphology. Once outliers were detected for each pattern, it was investigated if the outliers fell within a consistent area of the impact pattern on the target surface, as well as what potential causes may contribute to these outliers. Finally, it was hypothesized that an operator of the software used in this project might improve over time, and therefore this was tested. To investigate these aims, 55 patterns were created using a reproducible impact device which created large patterns with a high number of bloodstains. These patterns were then scanned using the FARO Focus3D S120 laser scanner, and processed using the accompanying FARO SCENE software with a Forensic Plug-in containing a BPA component. All stains fitting accepted stain criteria from each pattern were processed on the software, creating a calculated origin for each pattern. Each pattern was then examined for outliers by measuring which trajectories fell greater than 15 cm from the true origin. To investigate the fluid dynamics involved in origin determination error, a FASTCAM SA1.1 high speed camera was used to capture the blood droplets of 13 impact patterns during flight. The patterns were processed for outliers using the FARO SCENE software, and the footage analyzed to visualize the outliers during flight. Finally, the distance between the true and calculated origin for each of the first 55 patterns was measured to represent accuracy, and analyzed to determine if it decreased over time, representing an increase in accuracy due to operator experience. A distance of 15 cm from the true origin was decided upon as a limit for acceptable trajectories, and therefore any trajectories falling outside this limit were determined to be outliers. It was found that most of the patterns in this research produced outliers, and they were not recognizable by their stain morphology. The impact patterns were divided into 16 sections (4x4) and it was found that 3 of the middle sections (6, 7 and 11) produced a significantly higher number of outliers than other sections of the pattern. These outliers were found to skew to the right of the blood source. No distinct causes of outliers were determined; however the possibility of oscillating blood droplets, disintegrating blood droplets, as well as two overlapping stains could not be eliminated as possible causes of outlier trajectories. Finally, there was a significant increase in accuracy over time using the FARO SCENE software for the first 11 patterns; all following patterns showed a consistent level of accuracy. This research provides information regarding the error involved in area of origin determinations: outliers are consistently generated in the same general area of impact patterns, and skew to the right of the true origin. This could have a significant effect on stain selection in the future, as well as considerations when reporting area of origin, such as potential inherent error to the right of the true origin. It should be considered by BPA analysts in the future to optimize the accuracy of this discipline.