Proceedings for Annual Meeting of The Japanese Pharmacological Society
Online ISSN : 2435-4953
The 97th Annual Meeting of the Japanese Pharmacological Society
Session ID : 97_3-B-S57-1
Conference information

Symposium
Current status of automation of pharmacotoxicity testing using images and machine learning
*Takahisa MurataKoji KobayashiNaoaki SakamotoYusuke Miyazaki
Author information
CONFERENCE PROCEEDINGS OPEN ACCESS

Details
Abstract

Digital transformation (DX) is now being touted in various areas of drug discovery research. However, automation of non-clinical studies and basic research using laboratory animals is highly challenging. In our laboratory, we have been developing an animal behavior analysis system using images and machine learning for six years. We are now in the process of developing a technology that not only automates animal experiments, but also captures new animal phenotypes over long periods of time, day and night, that have not been visible to humans before, and reduces them to structural values. It is also very significant in that it allows us to evaluate animal phenotypes in an unbiased manner. Currently, it is possible to 1) identify individual mice and rats, 2) evaluate their spontaneous movements including eating and drinking, 3) identify and track tissue sites, 4) detect specific behaviors such as scratching, grooming, and standing up, and 5) detect facial expressions such as pain. At the same time, we are also working to generalize these technologies so that they can be used in your laboratories in the near future. In this symposium, we would like to introduce current technologies and future challenges. We hope that these technologies will contribute to animal welfare and provide a new impact for pharmacotoxicological research.

Content from these authors
© 2023 The Authors(s)
Previous article Next article
feedback
Top