Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Technical note : Automated data extraction from autopsy reports using a custom Python script

Busch, Johannes Rødbro and Wingren, Carl Johan LU (2026) In Forensic Science International 379.
Abstract

Evidence-based forensic research necessitates the creation of large and valid datasets. However, in our experience many departments face a challenge in how to extract this data from electronically archived records. This technical note describes a custom script created in the Python programming language. The program can extract data on decedent sex, age, body height, body weight, organ weight, organ dimensions, degree of putrefaction, listed cause of death, medical history and scene description from approximately 23,000 records in under two hours. Validity for many of these data are around 97-99 %. The program can be modified to extract any type of information. Data that are structured uniformly in the records result in higher data... (More)

Evidence-based forensic research necessitates the creation of large and valid datasets. However, in our experience many departments face a challenge in how to extract this data from electronically archived records. This technical note describes a custom script created in the Python programming language. The program can extract data on decedent sex, age, body height, body weight, organ weight, organ dimensions, degree of putrefaction, listed cause of death, medical history and scene description from approximately 23,000 records in under two hours. Validity for many of these data are around 97-99 %. The program can be modified to extract any type of information. Data that are structured uniformly in the records result in higher data validity. Compared with manual extraction of data, automated extraction provide several benefits, including speed, accuracy, and flexibility.

(Less)
Please use this url to cite or link to this publication:
author
and
publishing date
type
Contribution to journal
publication status
published
in
Forensic Science International
volume
379
article number
112756
publisher
Elsevier
external identifiers
  • pmid:41319600
ISSN
1872-6283
DOI
10.1016/j.forsciint.2025.112756
language
English
LU publication?
no
additional info
Copyright © 2025 The Authors. Published by Elsevier B.V. All rights reserved.
id
6f922324-9ebe-40bb-985f-e7e3f0af8a3c
date added to LUP
2025-12-01 10:44:16
date last changed
2025-12-01 10:44:16
@article{6f922324-9ebe-40bb-985f-e7e3f0af8a3c,
  abstract     = {{<p>Evidence-based forensic research necessitates the creation of large and valid datasets. However, in our experience many departments face a challenge in how to extract this data from electronically archived records. This technical note describes a custom script created in the Python programming language. The program can extract data on decedent sex, age, body height, body weight, organ weight, organ dimensions, degree of putrefaction, listed cause of death, medical history and scene description from approximately 23,000 records in under two hours. Validity for many of these data are around 97-99 %. The program can be modified to extract any type of information. Data that are structured uniformly in the records result in higher data validity. Compared with manual extraction of data, automated extraction provide several benefits, including speed, accuracy, and flexibility.</p>}},
  author       = {{Busch, Johannes Rødbro and Wingren, Carl Johan}},
  issn         = {{1872-6283}},
  language     = {{eng}},
  publisher    = {{Elsevier}},
  series       = {{Forensic Science International}},
  title        = {{Technical note : Automated data extraction from autopsy reports using a custom Python script}},
  url          = {{http://dx.doi.org/10.1016/j.forsciint.2025.112756}},
  doi          = {{10.1016/j.forsciint.2025.112756}},
  volume       = {{379}},
  year         = {{2026}},
}