SIGN IN/SIGN UP

Automation, Optimisation and Dissemination of the Mouse Grimace Scale

Project Objectives

  • Develop enhanced, standardised protocols for accurately applying the Mouse Grimace Scale across research facilities.
  • Design, build, and optimise an automated experimental setup capable of capturing mouse facial data for grimace scoring.
  • Create automated tools that support consistent, objective identification of facial action units.
  • Validate automated scoring outputs against expert human scoring to ensure reliability.
  • Reduce observer‑dependent variability by defining clearer criteria, improved training materials, and user guidance.
  • Facilitate uptake by producing accessible documentation, training resources, and dissemination activities.

How This Advances 3Rs Implementation

  • Supports broad implementation of refined pain‑assessment protocols by making them easier to use and less time‑intensive.
  • Strengthens reproducibility across laboratories, reducing the total animal numbers needed for studies.
  • Introduces automation that allows non‑specialist staff to conduct high‑quality welfare assessments, improving institutional capacity.
  • Encourages a cultural shift toward proactive, evidence‑based welfare monitoring.
  • Creates transferable workflows that can support refinement in other behavioural or physiological scoring systems.

Background

Pain assessment is essential for maintaining high welfare standards in research involving mice. Traditional behavioural scoring relies on trained humans manually identifying subtle facial changes, which can be subjective, inconsistent, and difficult to scale. The Mouse Grimace Scale is a well-established method to identify discomfort through facial action units, yet widespread adoption has been hindered by time consumption, variability between scorers, and limited automation.

This project addresses these challenges by developing improved protocols and automated scoring tools. By reducing user subjectivity, increasing standardisation, and lowering barriers for implementation, the project strengthens both scientific quality and the welfare of animals involved in research.

Published : 08.07.25

PROJECT DETAILS 

  

Grant scheme: Refinement Grant 

Grant number: RG-2022-010 

Status: Complete

Funding amount: CHF 25’000 

Animal use: License obtained

──────────────────── 

Start date: 01.02.23 

End date: 31.01.24

Dr Oliver Sturman

ETH Zürich

OUTPUT

 

The system and related expertise has been shared with collaborators at ETH Zurich, Bern, Münster, and Roche. It has also been showcased at several events, such as the Artificial Intelligence in Animal Welfare session at the 13th CABMM Symposium, the 64th SOT Annual Meeting and the Animal Pain and AI workshop in Bern during 2025.

The team is now focusing on preparing a manuscript for publication, a preprint for which can be found here:

Sturman, O., Schmutz, M., Lorimer, T., Zhang, R., Privitera, M., Roessler, F. K., Leonardi, J., Waag, R., Marinescu, A.-M., Bekemeier, C., Hohlbaum, K., & Bohacek, J. (2025). GrimACE: Automated, multimodal cage-side assessment of pain and well‑being in mice. bioRxiv. https://doi.org/10.1101/2025.03.07.642046.