How to solve murders using post-mortem metabolomic fingerprints and artificial intelligence

Postmortem diagnosis of several types of murders, poisonings, and more relies on subjective evaluations of findings or non-specific patterns, and these causes of death will thus mainly be established in an indirect way.

Such deaths include, but are not limited to, suffocation (asphyxia), anaphylactic shock and poisonings by insulin- and other atypical toxins.Postmortem metabolomics fingerprints consist of patterns of metabolites reflecting biochemical processes in different tissues. Such fingerprints reveal the agonal phase, i.e. events preceding death.

It also reflects the cause and the time of death and could significantly improve postmortem diagnoses.

However, to identify metabolomic fingerprints for specific agonal and post-mortem events, we would need to combine scientific competence of forensic pathologists/toxicologists, analytical chemists, and machine learning/artificial intelligence experts.The overall aim of the proposal is to create a research environment that brings these experts together.

The research environment includes projects for postmortem metabolomic fingerprints and mathematical models that can identify 1) insulin, cyanide, heavy metal, and herbal poisonings; 2) asphyxia, anaphylactic shock; and 3) time of death by combining post-mortem interval and circadian rhythms.A successful project will aid police work in giving them investigative leads and new tools to find the cause of death in cases where we today lack scientific and objective forensic evidence.