Related Experiment Video
Updated: Aug 6, 2026
05:33
Introduction of an Integrated Pathology Image Management, Artificial Intelligence, and Reporting System
Published on: July 11, 2025
Artificial intelligence for short-term post-mortem interval estimation: a systematic review with PROBAST and
Gianmarco Sirago1, Biagio Solarino1, Alessandro Dell'Erba1
1Section of Legal Medicine, University of Bari Aldo Moro, Piazza Giulio Cesare 11, 70124, Italy.
Legal Medicine (Tokyo, Japan)
|July 18, 2026
Summary
Artificial intelligence and machine learning (AI/ML) show promise for estimating post-mortem interval (PMI). However, current human studies for short-term PMI estimation are limited by small datasets and incomplete reporting, hindering reliability.
Area of Science:
- Forensic Science
- Biomedical Informatics
- Artificial Intelligence
Background:
- Estimating post-mortem interval (PMI) is crucial in forensic casework.
- Traditional PMI estimation methods lack precision and are context-dependent.
- Artificial intelligence and machine learning (AI/ML) offer potential for improved PMI estimation using diverse data.
Purpose of the Study:
- To systematically review human studies developing or validating AI/ML models for short-term PMI estimation (≤72 hours).
- To assess the methodological rigor, risk of bias, and reporting completeness of AI/ML models for PMI.
- To identify limitations and guide future research in AI/ML-based PMI estimation.
Main Methods:
- Systematic literature search across major scientific databases (PubMed, Web of Science, Scopus, IEEE Xplore, Cochrane Library) up to December 19, 2025.
- Data extraction using CHARMS (Consensus Health Assessment and Reporting of Medical Studies).
- Risk of bias and applicability assessed using PROBAST (Probabilistic Robustness of Diagnostic Test Accuracy Studies).
- Reporting completeness evaluated with TRIPOD+AI (Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis Or Diagnosis).
Main Results:
- Only 8 human studies met stringent inclusion criteria for short-term PMI estimation.
- Diverse AI/ML approaches were evaluated, including corneal opacity imaging, vitreous biochemistry, ATR-FTIR spectroscopy, post-mortem CT radiomics, and microbiomics.
- Studies frequently featured small, single-center datasets, high risk of bias (PROBAST), and common reporting gaps (TRIPOD+AI), particularly regarding sample size, missing data, calibration, and model availability.
Conclusions:
- AI/ML methods for PMI estimation show promise but are currently constrained by limited human evidence.
- Small, heterogeneous datasets, insufficient transportability testing, and incomplete reporting impede the reliability and generalizability of current AI/ML models.
- Future research necessitates adequately powered, multi-center studies with robust validation, calibration assessment, and transparent, open reporting aligned with TRIPOD+AI guidelines.