Scribd
Upload
46 views

Bloodstain Pattern Analysis: Presented By: Dr. Hal Stern

Bloodstain Pattern Analysis was presented by Dr. Hal Stern. The presentation summarized accomplishments from the CSAFE 1.0 project including publishing two freely available bloodstain pattern datasets and developing frameworks for determining the region of origin and classifying patterns. The goals of the CSAFE 2.0 project are to develop objective analytic approaches to enhance bloodstain pattern analysis through image processing and statistical methods for assessing causal mechanisms.

Uploaded by

ritvikr
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
46 views

Bloodstain Pattern Analysis: Presented By: Dr. Hal Stern

Bloodstain Pattern Analysis was presented by Dr. Hal Stern. The presentation summarized accomplishments from the CSAFE 1.0 project including publishing two freely available bloodstain pattern datasets and developing frameworks for determining the region of origin and classifying patterns. The goals of the CSAFE 2.0 project are to develop objective analytic approaches to enhance bloodstain pattern analysis through image processing and statistical methods for assessing causal mechanisms.

Uploaded by

ritvikr
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 9

Bloodstain Pattern Analysis

Presented by: Dr. Hal Stern

May 13, 2020 forensicstats.org


CSAFE 1.0 Accomplishments

Major Accomplishments:
• Published two freely available datasets (impact bloodspatters) available at forensicstats.org/data
• Attinger et al., 2018, Data in Brief – Impact beating spatters
• Attinger et al., 2019, Data in Brief – Gunshot backspatters
• Framework for determining the region of origin of blood spatters considering fluid dynamics and statistical uncertainties
(Attinger et al., 2019, Forensic Science International)
• Framework to automatically classify bloodstain patterns caused by gunshot and blunt impact at various distances
(Liu et al., 2020, Journal of Forensic Sciences)
• Image processing pipeline to approximate bloodstain patterns as a set of elliptical-shaped stains (Zou et al., submitted
manuscript)
• Preliminary work developing videos of blood spatter formation and statistical tools for analysis of the videos
(manuscript in progress)

Impact:
• The high-resolution blood spatter database provides a consistent set of data using controlled and carefully documented
experiments available to researchers worldwide
• First work on determine of region of origin with quantification of uncertainty
• Computer based decision process for distinguishing bloodstain patterns caused by blunt or gunshot impacts

forensicstats.org | 2
Data
Bloodstain patterns generated by blunt impact & gunshots
(Attinger et al. 2018, 2019 in Data in Brief)

forensicstats.org | 3
Classification of bloodstain patterns

• Application of data
• Distinguishing mechanisms
• Examine distance of stains
around center

centroid

forensicstats.org | 4
Experimental system to capture
high-speed video of bloodspatter

• High-speed cameras to capture video of blood droplet


• Vary droplet size, angle of inclined plane
• Record size and shape of droplet over time
• Repeated measurements from same settings provide
information about variability

forensicstats.org | 5
Research Area Objectives

• Two main questions arise with BPA: determination of mechanism and origin.

• There has been limited analysis of the accuracy of bloodstain pattern analysts in
determining the causal mechanism for a particular bloodstain pattern.

• Laber et al. (2014) describe the results of several studies and find many inconclusive
opinions, incorrect mechanisms identified in 10% or more of cases, and the dependence
of analyst performance on the context in which they carried out the analysis.

• Goal of CSAFE 2.0 research project: Develop objective analytic approaches to enhance
the practice of BPA.
• Analysis tools for bloodstain pattern images
• Definition and refinement of relevant features to distinguish mechanisms
• Methods and models for probabilistic analysis of mechanism and origin

forensicstats.org | 6
forensicstats.org | 6
CSAFE 2.0 Bloodstain Projects and
Lead Investigators

BPA I- Statistical Methods for Bloodstain Pattern Analysis


Lead PI: Hal Stern, UCI

Other participants:
Kris de Brabanter, Iowa State
Tong Zou, UCI (graduate student)
Michael Taylor, EZR (New Zealand) (collaborator)

forensicstats.org | 7
CSAFE 2.0

Proposed Activities:
• Image processing pipeline
• Starts with a bloodstain pattern image
• Approximates it by a collection of one or more ovoid/elliptical shapes
• Develop features from these approximations
• Develop score-based or feature-based likelihood ratio for assessing alternative
hypotheses about the mechanism causing the pattern.
• Builds on the CSAFE 1.0 work of the Iowa State group (De Brabanter)
• Collaborate with practitioners to assess the practicality of the approach in casework.

Potential Impact:
• Current evaluation of mechanism for bloodstain pattern analysis is subjective.
• First steps in assessing a likelihood ratio approach to bloodstain pattern analysis.
forensicstats.org | 8
Resources and Needs

• Ways forensic science community can engage/partner with us:


• All of our data are laboratory generated; hopeful that as methods develop we can explore
their potential in case-like (or case) materials

forensicstats.org | 9

You might also like