Applied Research, Development, and Method Validation for a Statistically Based Comparison of Tool Marks using GelSight-Based 3D Imaging and Novel Comparison Algorithms for Firearm Forensics

As submitted by the proposer: The proposed work represents a critical next step in investigating the overall performance characteristics of a novel system for structural 3D imaging and comparison of cartridge casings, TopMatch. In January 2013, we began development of a 3D surface topography imaging and analysis system for firearm forensics based on the GelSight imaging technology and custom feature-based image comparison algorithms. The technology shows excellent match accuracy. However, to fully establish the base credibility of the platform it is necessary to identify best scanning practices and demonstrate that the method meets the quality control criteria of other forensic instruments. This proposal aims to greatly enhance this novel, low-cost system for structural 3D imaging and comparison of cartridge casings by establishing best practices and conducting a set of methodology studies. The aims of this proposal follow the intent of the Scientific Working Group on DNA Analysis Methods Validation Guidelines for Analysis Methods document. The report describes validation criteria for new scientific methods. We propose to complete a series of experiments to address the issues of Precision, Repeatability, Reproducibility, Contamination Risk, and Performance Checks. In Aim 1 we will setup three datasets and develop objective measures of scan quality. In Aim 2 we will establish a set of Best Scanning Practices, explore Cleaning Protocols, Focus Variation, and Ambient Lighting effects, investigate Precision, Repeatability, and Contamination Risk, and establish Reproducibility via an Inter-Operator Validation study. In Aim 3 we will develop System Performance Checks. Several completed feasibility studies described in this proposal, support the presented aims. The proposal will be completed by Cadre Research Labs, a scientific computing contract research organization, working in collaboration with GelSight Inc, a company formed by the MIT researchers who developed the GelSight surface topography imaging technology. The two companies will collaborate with ToddWeller, a firearms identification specialist and Criminalist in the Oakland Police Department. We will continue collaborations with colleagues at the San Francisco PD, the Contra Costa County Sheriffs Office, the U. of Illinois at Chicago, and NIST. The proposed project will establish best scanning practices and fully validate the operational characteristics of the novel 3D imaging and analysis system. This will benefit law enforcement and their ability to present firearm identification and tool mark evidence in the courtroom. Interim reports will be provided as required by the NIJ. A final report will summarize the completion and results of the three aims. ca/ncf