Postmortem iris recognition and its application in human identification
MetadataShow full item record
Iris recognition is a validated and non-invasive human identification technology currently implemented for the purposes of surveillance and security (i.e. border control, schools, military). Similar to deoxyribonucleic acid (DNA), irises are a highly individualizing component of the human body. Based on a lack of genetic penetrance, irises are unique between an individual’s left and right iris and between identical twins, proving to be more individualizing than DNA. At this time, little to no research has been conducted on the use of postmortem iris scanning as a biometric measurement of identification. The purpose of this pilot study is to explore the use of iris recognition as a tool for postmortem identification. Objectives of the study include determining whether current iris recognition technology can locate and detect iris codes in postmortem globes, and if iris scans collected at different postmortem time intervals can be identified as the same iris initially enrolled. Data from 43 decedents involving 148 subsequent iris scans demonstrated a subsequent match rate of approximately 80%, supporting the theory that iris recognition technology is capable of detecting and identifying an individual’s iris code in a postmortem setting. A chi-square test of independence showed no significant difference between match outcomes and the globe scanned (left vs. right), and gender had no bearing on the match outcome. There was a significant relationship between iris color and match outcome, with blue/gray eyes yielding a lower match rate (59%) compared to brown (82%) or green/hazel eyes (88%), however, the sample size of blue/gray eyes in this study was not large enough to draw a meaningful conclusion. An isolated case involving an antemortem initial scan collected from an individual on life support yielded an accurate identification (match) with a subsequent scan captured at approximately 10 hours postmortem. Falsely rejected subsequent iris scans or "no match" results occurred in about 20% of scans; they were observed at each PMI range and varied from 19-30%. The false reject rate is too high to reliably establish non-identity when used alone and ideally would be significantly lower prior to implementation in a forensic setting; however, a "no match" could be confirmed using another method. Importantly, the data showed a false match rate or false accept rate (FAR) of zero, a result consistent with previous iris recognition studies in living individuals. The preliminary results of this pilot study demonstrate a plausible role for iris recognition in postmortem human identification. Implementation of a universal iris recognition database would benefit the medicolegal death investigation and forensic pathology communities, and has potential applications to other situations such as missing persons and human trafficking cases.