Howard Theodore Harcke MD, Presenter: Nothing to Disclose

Reed Selwyn PhD, Abstract Co-Author: Nothing to Disclose

Charles Allen, Abstract Co-Author: Nothing to Disclose

Edward L. Mazuchowski MD, PhD, Abstract Co-Author: Nothing to Disclose

A logical forensic application of postmortem CT is the characterization of metal fragments present in the body at the time of death investigation. The purpose of this study is to evaluate the application of standard region of interest (ROI) analysis to extended CT scale measurements of density (Hounsfield Units, HU). The goals were (1) to determine if fragments in a body are of the same material and (2) to predict the composition of metal fragments based on density relative to samples of known composition.

Postmortem CT imaging was performed on a GE Lightspeed 16x scanner with extended scale. Analysis of metal fragments was made on a GE Advantage Workstation 4.2 using the standard software tools for measuring HU’s from regions of interest. Variations in ROI placement and size within fragments were studied. Density measurements were made in the body and in air after recovery at autopsy. Sample objects of known composition (steel, copper, brass, and aluminum) were scanned in water and air for comparison and effects of kVp and mA were evaluated.

Both recovered fragments and fragments of known composition showed highly variable estimates of density. Tested fragments ranged from 1200 to 30,000 HU. Values were dependent on ROI size and location both longitudinally and transaxially. Measurements in the body produced densities approximately 50% lower than values measured in air. Lower scan kVp produced higher HU values in both steel and aluminum (air measurement). Because of the wide variation in extended scale HU measurements occurring within and between fragments it was not possible to conclude whether fragments were from the same source material or to distinguish composition of higher density fragments (steel, copper, brass).

Scatter, beam hardening and ROI analysis affect the CT measurement of metal fragments in the body to such a degree that extended scale determination of density should not be used in forensic analysis without further investigation.

As postmortem CT is applied to death investigation, standards to determine metal fragment characteristics need to be defined. New or revised algorithms should be considered.

Harcke, H, Selwyn, R, Allen, C, Mazuchowski, E, Postmortem Computed Tomography (CT) Analysis of Metal Fragments: Forensic Limitations.  Radiological Society of North America 2012 Scientific Assembly and Annual Meeting, November 25 - November 30, 2012 ,Chicago IL. http://archive.rsna.org/2012/12025362.html