New AI tool predicts post-mortem time with precision
AI analysis of blood metabolites can now estimate time of death with about one-day accuracy, improving forensic investigations and police work.
Researchers at Linköping University and the Swedish National Board of Forensic Medicine have developed an AI tool that estimates time of death from blood metabolites. The model, trained on thousands of samples, provides greater accuracy than traditional forensic methods.
Methods like body temperature, rigor mortis, or eye potassium become unreliable after a few days. AI analysis of blood metabolites estimates time of death with about one-day accuracy for up to 13 days post-mortem.
The project uses a unique data resource of over 45,000 autopsies, with 4,876 samples used to train the AI. Researchers say the method works globally, even in labs with smaller datasets, making it useful for forensic investigations.
Next steps aim to increase precision, allowing models to estimate not only the day but also the specific time of death. Experts say the tool can improve investigations by guiding law enforcement and aiding complex cases.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
