50th MEETING

European Brain and Behaviour Society

Amsterdam, Netherlands · 26-29 August 2023

50th MEETING

European Brain and Behaviour Society

Amsterdam, Netherlands
26-29 August 2023

50th MEETING

European Brain and Behaviour Society

Amsterdam, Netherlands
26-29 August 2023

EBBS Satellite Event

26 August 2023

A new outlook on automated analysis of animal behavior: implementing a machine learning-based phenotyping

This event is open to a limited number of students and postocs that participate also in the EBBS Meeting. Slots are available on a first-come, first-served basis. The cost is 25 EUR, to be paid with the meeting registration fee. During the registration procedure, please indicate your intention to participate in this satellite event.

Learning goals:

  • How to use automated home-cage monitoring technologies in rodents? How can they help in my studies?
  • How to analyze the data? What are the challenges, and what are the solutions?
  • Hands-on workshop on how to use DeepLabCut to analyze your data.

1. Home cage-systems: let’s start at the beginning (presentation, about 60 min)

Lior Bikovski, PhD, Director of the Myers Neuro-Behavioral Core Facility, of the SICF group Sackler Faculty of Medicine Tel Aviv University; TeaTime COST Action.

Reproducibility and replicability are crucial in all fields of experimental research, but even more so in animal research, where the lives and welfare of the animals are valuable for ethical reasons, and should not be wasted for inconclusive research. A problem inherent to the field of phenotyping is that the environment (e.g. the cage, the lab, the human handler) significantly affects the final phenotyping process of our mice and rats. In an attempt to reduce the external effects and to create a standardized environment, we performed phenotyping in automated home-cage monitoring systems. These systems offer a fixed environment, enable prolonged and longitudinal recordings, and provide large continuous measurements of spontaneous “natural” behavior that can be analyzed across multiple time scales. Selecting the optimal system for each scientific project requires knowledge of the different systems that are commercially available, and understanding what features are needed. In this talk, I will present TeaTime COST Action, a networking group dedicated to disseminating the knowledge about home-cage systems. In addition, I will review the basic features of a “home-cage” and discuss the different technologies available today.

2. How to use DeepLabCut to analyze your data? (workshop, about 120 min)

Konrad Danielewski, Laboratory of Emotions Neurobiology, BRAINCITY – Centre of Excellence for Neural Plasticity and Brain Disorders, Nencki Institute of Experimental Biology, Warsaw, Poland.

DeepLabCut is a toolbox for markerless pose estimation of animals performing various tasks. It allows the user to track arbitrarily chosen body parts across a video of one or multiple animals. DeepLabCut has a very straightforward workflow that can be followed by calling specific high-level functions or an intuitive GUI. This workshop will focus on introducing key concepts behind pose estimation and leveraging the possibilities it provides in making behavioral analysis more robust, in-depth, and streamlined. We will talk about the following:

  1. What is pose estimation? How does it work?
  2. What is behavior, and how to quantify it?
  3. Hands-on training to start your DeepLabCut journey*.

*The participants will need their laptops with pre-installed software; we will send all necessary links before the workshop. We will provide access to the learning material to those that have attended.